The isotopic signature of U^V during bacterial reduction

A.R. Brown¹,

¹École Polytechnique Fédérale de Lausanne (EPFL), Environmental Microbiology Laboratory, CH-1015 Lausanne, Switzerland

M. Molinas¹,

¹École Polytechnique Fédérale de Lausanne (EPFL), Environmental Microbiology Laboratory, CH-1015 Lausanne, Switzerland

Y. Roebbert²,

²Institute of Mineralogy, Leibniz University Hannover, D-30167 Hannover, Germany

R. Faizova³,

³École Polytechnique Fédérale de Lausanne (EPFL), Group of Coordination Chemistry, CH-1015 Lausanne, Switzerland

T. Vitova⁴,

⁴Karlsruhe Institute of Technology (KIT), Institute for Nuclear Waste Disposal (INE), D-76021 Karlsruhe, Germany

A. Sato^5,6,

⁵Department of Chemistry, Tokyo Metropolitan University, Tokyo, Japan ⁶Department of Chemistry, Hiroshima University, Hiroshima, Japan

M. Hada⁵,

⁵Department of Chemistry, Tokyo Metropolitan University, Tokyo, Japan

M. Abe^5,6,

⁵Department of Chemistry, Tokyo Metropolitan University, Tokyo, Japan ⁶Department of Chemistry, Hiroshima University, Hiroshima, Japan

M. Mazzanti³,

³École Polytechnique Fédérale de Lausanne (EPFL), Group of Coordination Chemistry, CH-1015 Lausanne, Switzerland

S. Weyer²,

²Institute of Mineralogy, Leibniz University Hannover, D-30167 Hannover, Germany

R. Bernier-Latmani¹

¹École Polytechnique Fédérale de Lausanne (EPFL), Environmental Microbiology Laboratory, CH-1015 Lausanne, Switzerland

Affiliations | Corresponding Author | Cite as | Funding information

Geochemical Perspectives Letters v29 | https://doi.org/10.7185/geochemlet.2411
Received 13 July 2023 | Accepted 1 March 2024 | Published 9 April 2024

Keywords: Isotope fractionation, pentavalent uranium, U^V, microbial U^VI reduction



PDF	PDF+SI

Share this article
Article views:
523
Cumulative count of HTML views and PDF downloads.
Download Citation
Rights & Permissions

top

Abstract

The two step electron transfer during bacterial reduction of U^VI to U^IV is typically accompanied by mass-independent fractionation of the ²³⁸U and ²³⁵U isotopes, whereby the heavy isotope accumulates in the reduced product. However, the role of the U^V intermediate in the fractionation mechanism is unresolved due to the challenges associated with its chemical stability. Here, we employed the U^V stabilising ligand, dpaea^2-, to trap aqueous U^V during U^VI reduction by Shewanella oneidensis. Whilst the first reduction step from U^VI to U^V displayed negligible fractionation, reduction of U^V to U^IV revealed mass-dependent isotope fractionation (preferential reduction of the ²³⁵U), contrary to most previous observations. This surprising behaviour highlights the control that the U-coordinating ligand exerts over the balance between reactant U supply, electron transfer rate, and U^IV product sequestration, suggesting that U^V speciation should be considered when using U isotope ratios to reconstruct environmental redox conditions.

Figures

Figure 1 (a) Uranium mass distribution in sacrificial reactors containing S. oneidensis incubated with U^VIO₂-dpaea. (b) Normalised U M₄-edge HR-XANES spectrum of aqueous uranium after 144 hr of incubation with S. oneidensis, along with U^VIO₂-dpaea, U^VO₂-dpaea^– and U^IV-(dpaea)₂ standards.	Figure 2 (a) Aqueous uranium concentrations throughout the first 24 hr of incubation of U^VIO₂-dpaea and S. oneidensis. Symbols and error bars depict one standard deviation of the mean of duplicate reactors. (b) Corresponding δ²³⁸U values of the aqueous U in duplicate systems (A and B), reported as a fraction of the maximum aqueous U concentration. Symbols and error bars depict two standard deviations of the mean of triplicate measurements. The δ²³⁸U value of the initial U^VIO₂-dpaea is plotted as a yellow dotted line. (c) Aqueous uranium concentrations throughout the whole reaction between U^VIO₂-dpaea and S. oneidensis. Symbols and error bars depict one standard deviation of the mean of duplicate reactors. (d) δ²³⁸U values of the aqueous U after 24 hr when the aqueous U concentration began to decrease. Values are reported as a fraction of the maximum aqueous U concentration. Symbols and error bars depict two standard deviations of the mean of triplicate measurements. The Rayleigh model (blue dashed line) corresponds to the linear best fit of the logarithmic data, R² = 0.89, from which the isotope enrichment factor, ɛ, is derived.	Figure 3 (a) Aqueous uranium concentrations during equilibrium isotope exchange experiments between U^VO₂-dpaea^– with an initial isotopic composition of ∼5 ‰, and U^IV present as the product of the bioreduction experiments of natural U, with an initial isotopic composition of 0 ‰. Symbols and error bars depict one standard deviation of the mean of duplicate reactors. (b) δ²³⁸U values of the aqueous U. Symbols and error bars depict one standard deviation of the mean of duplicate reactors.	Figure 4 Cartoon of the proposed mechanism of U isotope reduction and fractionation for both U^VI-carbonate (left) and U^V-dpaea^– (right). Electrons are transferred from the cell to outer membrane U reducing proteins (blue areas) containing multiple redox active heme iron centres (red circles). Depending on the flux of electrons, the heme iron centres are either in their reduced state (solid fill) or oxidised state (open fill).
Figure 1	Figure 2	Figure 3	Figure 4

View all figures and tables

top

Introduction

Hexavalent uranium (U^VI) is the predominant oxidation state of U under ambient oxic conditions at Earth’s surface and forms soluble uranyl complexes. Under anoxic conditions, reduction of U^VI to tetravalent U (U^IV) can be mediated by an array of microorganisms or abiotically via Fe(II)- or sulfide-bearing compounds (Basu et al., 2014

Basu, A., Sanford, R.A., Johnson, T.M., Lundstrom, C.C., Löffler, F.E. (2014) Uranium isotopic fractionation factors during U(VI) reduction by bacterial isolates. Geochimica et Cosmochimica Acta 136, 100–113. https://doi.org/10.1016/j.gca.2014.02.041

; Brown et al., 2018

Brown, S.T., Basu, A., Ding, X., Christensen, J.N., DePaolo, D.J. (2018) Uranium isotope fractionation by abiotic reductive precipitation. Proceedings of the National Academy of Sciences of the United States of America 115, 8688–8693. https://doi.org/10.1073/pnas.1805234115

), resulting in the precipitation of sparingly soluble U^IV species. This behaviour has been harnessed for the (bio)remediation of U contaminated groundwater.Such U redox transformations are often accompanied by changes in the ²³⁸U/²³⁵U ratio, reported as δ²³⁸U (Andersen et al., 2017

Andersen, M.B., Stirling, C.H., Weyer, S. (2017) Uranium Isotope Fractionation. Reviews in Mineralogy and Geochemistry 82, 799–850. https://doi.org/10.2138/rmg.2017.82.19

). Both ab initio calculations and isotope exchange experiments indicate that, at equilibrium, heavy ²³⁸U is enriched in the U^IV oxidation state (Schauble, 2007

Schauble, E.A. (2007) Role of nuclear volume in driving equilibrium stable isotope fractionation of mercury, thallium, and other very heavy elements. Geochimica et Cosmochimica Acta 71, 2170–2189. https://doi.org/10.1016/j.gca.2007.02.004

; Abe et al., 2008

Abe, M., Suzuki, T., Fujii, Y., Hada, M., Hirao, K. (2008) An ab initio molecular orbital study of the nuclear volume effects in uranium isotope fractionations. The Journal of Chemical Physics 129, 164309. https://doi.org/10.1063/1.2992616

; Wang et al., 2015

Wang, X., Johnson, T.M., Lundstrom, C.C. (2015) Low temperature equilibrium isotope fractionation and isotope exchange kinetics between U(IV) and U(VI). Geochimica et Cosmochimica Acta 158, 262–275. https://doi.org/10.1016/j.gca.2015.03.006

). This mass-independent fractionation arises from the nuclear field shift effect (NFSE), due to differences in the size and shape of the nuclei of heavy element isotopologues (Bigeleisen, 1996

Bigeleisen, J. (1996) Nuclear size and shape effects in chemical reactions. Isotope chemistry of the heavy elements. Journal of the American Chemical Society 118, 3676–3680. https://doi.org/10.1021/ja954076k

; Schauble, 2007

). At equilibrium, the NFSE is larger than, and operates in the opposite direction to, the conventional mass-dependent isotope effect, whereby the vibrational zero point energy of the lighter isotope leads to its enrichment in U^IV as mass-dependent fractionation (MDF) (Bigeleisen, 1996

; Schauble, 2007

; Fujii et al., 2009

Fujii, T., Moynier, F., Albarède, F. (2009) The nuclear field shift effect in chemical exchange reactions. Chemical Geology 267, 139–156. https://doi.org/10.1016/j.chemgeo.2009.06.015

). Thus, enrichment of ²³⁸U in U^IV following U^VI reduction has also been attributed to a dominant NFSE (Weyer et al., 2008

Weyer, S., Anbar, A.D., Gerdes, A., Gordon, G.W., Algeo, T.J., Boyle, E.A. (2008) Natural fractionation of ²³⁸U/²³⁵U. Geochimica et Cosmochimica Acta 72, 345–359. https://doi.org/10.1016/j.gca.2007.11.012

; Basu et al., 2014

, 2020

Basu, A., Wanner, C., Johnson, T.M., Lundstrom, C., Sanford, R.A., Sonnenthal, E., Boyanov, M.I., Kemner, K.M. (2020) Microbial U isotope fractionation depends on U(VI) reduction rate. Environmental Science and Technology 54, 2295–2303. https://doi.org/10.1021/acs.est.9b05935

; Stirling et al., 2015

Stirling, C.H., Andersen, M.B., Warthmann, R., Halliday, A.N. (2015) Isotope fractionation of ²³⁸U and ²³⁵U during biologically-mediated uranium reduction. Geochimica et Cosmochimica Acta 163, 200–218. https://doi.org/10.1016/j.gca.2015.03.017

; Stylo et al., 2015

Stylo, M., Neubert, N., Wang, Y., Monga, N., Romaniello, S.J., Weyer, S., Bernier-Latmani, R. (2015) Uranium isotopes fingerprint biotic reduction. Proceedings of the National Academy of Sciences of the United States of America 112, 5619–24. https://doi.org/10.1073/pnas.1421841112

), despite not necessarily representing isotopic equilibrium conditions.As U isotope fractionation is predominantly associated with redox transformations, U isotope signatures have been utilised as a (1) monitoring tool tuned specifically to the reductive rather than adsorptive removal of U^VI during remediation (Bopp et al., 2010

Bopp, C.J., Lundstrom, C.C., Johnson, T.M., Sandford, R.A., Long, P.E., Williams, K.H. (2010) Uranium ²³⁸U/²³⁵U isotope ratios as indicators of reduction: Results from an in situ biostimulation experiment at Rifle, Colorado, U.S.A. Environmental Science and Technology 44, 5927–5933. https://doi.org/10.1021/es100643v

), and (2) palaeo-redox proxy, whereby the preferential reduction of ²³⁸U during marine anoxia is recorded in sedimentary rocks and can be used to reconstruct the pervasiveness of anoxia in past global oceans (Montoya-Pino et al., 2010

Montoya-Pino, C., Anbar, A.D., van de Schootbrugge, B., Oschmann, W., Pross, J., Arz, H.W., Weyer, S. (2010) Global enhancement of ocean anoxia during Oceanic Anoxic Event 2: A quantitative approach using U isotopes. Geology 38, 315–318. https://doi.org/10.1130/G30652.1

; Brennecka et al., 2011

Brennecka, G.A., Herrmann, A.D., Algeo, T.J., Anbar, A.D. (2011) Rapid expansion of oceanic anoxia immediately before the end-Permian mass extinction. Proceedings of the National Academy of Sciences of the United States of America 108, 17631–17634. https://doi.org/10.1073/pnas.1106039108

; Andersen et al., 2017

Andersen, M.B., Stirling, C.H., Weyer, S. (2017) Uranium Isotope Fractionation. Reviews in Mineralogy and Geochemistry 82, 799–850. https://doi.org/10.2138/rmg.2017.82.19

). Hence, it is crucial to constrain the mechanistic underpinnings of U isotope fractionation to improve the reliability of U isotope based redox reconstructions.One important aspect of the U reduction mechanism is the role of the pentavalent U (U^V) intermediate. Previous studies have focused on the complete reduction of U^VI to U^IV. However, there is increasing evidence of the stabilisation and persistence of U^V intermediates within abiotic and biological systems (Roberts et al., 2017

Roberts, H.E., Morris, K., Law, G.T.W., Mosselmans, J.F.W., Bots, P., Kvashnina, K., Shaw, S. (2017) Uranium(V) incorporation mechanisms and stability in Fe(II)/Fe(III) (oxyhydr)oxides. Environmental Science and Technology Letters 4, 421–426. https://doi.org/10.1021/acs.estlett.7b00348

; Pan et al., 2020

Pan, Z., Bártová, B., Lagrange, T., Butorin, S.M., Hyatt, N.C., Stennett, M.C., Kvashnina, K.O., Bernier-latmani, R. (2020) Nanoscale mechanism of UO₂ formation through uranium reduction by magnetite. Nature Communications 11, 1–12. https://doi.org/10.1038/s41467-020-17795-0

).During microbiological U^VI reduction, two distinct mechanisms for the complete reduction to U^IV can occur: either via disproportionation of two uranyl^V atoms (generating U^VI and U^IV) (Vettese et al., 2020

Vettese, G.F., Morris, K., Natrajan, L.S., Shaw, S., Vitova, T., Galanzew, J., Jones, D.L. and Lloyd, J.R. (2020) Multiple lines of evidence identify U(V) as a key intermediate during U(VI) reduction by Shewanella oneidensis MR1. Environmental Science and Technology 54, 2268–2276. https://doi.org/10.1021/acs.est.9b05285

), or via a second biologically mediated electron transfer to U^V (Molinas et al., 2021

Molinas, M., Faizova, R., Brown, A., Galanzew, J., Schacherl, B., Bartova, B., Meibom, K.L., Vitova, T., Mazzanti, M. and Bernier-Latmani, R. (2021) Biological reduction of a U(V)-organic ligand complex. Environmental Science and Technology 55, 4753–4761. https://doi.org/10.1021/acs.est.0c06633

, 2023

Molinas, M., Meibom, K.L., Faizova, R., Mazzanti, M., Bernier-Latmani, R. (2023) Mechanism of reduction of aqueous U(V)-dpaea and solid-phase U(VI)-dpaea complexes: The role of multiheme c-type cytochromes. Environmental Science and Technology 57, 7537–7546. https://doi.org/10.1021/acs.est.3c00666

). However, due to the challenges associated with the chemical stabilisation and separation of U^V, there is a lack of experimental evidence for its isotopic fractionation, and thus its role in the fractionation mechanism remains unresolved.Ab initio calculations of the equilibrium isotope fractionation factor combined with a multi-step model of biological U^VI-carbonate reduction suggests that fractionation factors of up to 1.6 ‰ for the U^VI to U^V step and ∼0.8 ‰ for the U^V to U^IV step (a total of ∼2.4 ‰) may be expected (Sato et al., 2021

Sato, A., Bernier-Latmani, R., Hada, M., Abe, M. (2021) Ab initio and steady-state models for uranium isotope fractionation in multi-step biotic and abiotic reduction. Geochimica et Cosmochimica Acta 307, 212–227. https://doi.org/10.1016/j.gca.2021.05.044

). However, these values are significantly larger than those observed in nature or experimentally for U^VI to U^IV reduction, and it is not clear whether and how redox transformations to and from the U^V intermediate are involved in this discrepancy.The aminocarboxylate ligand dpaea²⁻ (dpaeaH₂ = bis(pyridyl-6-methyl-2-carboxylate)-ethylamine) can be used to precipitate both U^VI and U^IV whilst maintaining U^V as an aqueous complex at circumneutral pH (Faizova et al., 2018

Faizova, R., Scopelliti, R., Chauvin, A.-S., Mazzanti, M. (2018) Synthesis and characterization of a water stable uranyl(V) complex. Journal of the American Chemical Society 140, 13554–13557. https://doi.org/10.1021/jacs.8b07885

). These properties have allowed the reduction of U^VI by Shewanella oneidensis to be followed, revealing the potential for the biological reduction of the U^V intermediate, rather than its disproportionation (Molinas et al., 2021

, 2023

).Here, we leveraged the characteristics of dpaea²⁻ to trap aqueous U^V and provide direct experimental evidence of the U^V isotope signature during biological reduction by S. oneidensis. The observed isotopic fractionation factors were then compared to those predicted for equilibrium both computationally, using ab initio calculations, and experimentally, using isotope exchange approaches (see Supplementary Information for details). top

Results and Discussion

The overall experimental flow entails the biological reduction of U^VI-dpaea to first U^V-dpaea and then of U^V-dpaea to U^IV-dpaea. The temporal separation of the two steps, made possible by the vastly different reduction rates, allows the investigation of the isotopic fractionation of one step and then the other. Additionally, the equilibrium isotope fractionation factor was calculated via ab initio calculations. Finally, to investigate the equilibrium isotopic fractionation of U^V-dpaea and U^IV-dpaea, a heavy U^V-dpaea was incubated with a light U^IV-dpaea and the isotopic exchange probed over time.First, U^VI-dpaea was produced and reduced biologically. We incubated S. oneidensis with solid phase U^VIO₂-dpaea and observed a rapid decrease in U^VI over 24 hr. This was concomitant with an increase in aqueous U (Fig. 1a) comprising predominantly U^V (Fig. 1b) that was not observed in abiotic controls (Fig. S-1). Acidification of the aqueous U in 4.5 N HCl, in preparation for ion exchange chromatography, led to the detection of approximately equal quantities of U^VI and U^IV after separation (Fig. S-2), indicative of U^V disproportionation in the acidified preparation. Collectively, these data suggest that the first electron transfer was achieved rapidly, leading to the accumulation of U^V in solution, in agreement with previous studies (Molinas et al., 2021

, 2023

**Figure 1** **(a)** Uranium mass distribution in sacrificial reactors containing *S. oneidensis* incubated with U^VIO₂-dpaea. **(b)** Normalised U M₄-edge HR-XANES spectrum of aqueous uranium after 144 hr of incubation with *S. oneidensis*, along with U^VIO₂-dpaea, U^VO₂-dpaea^– and U^IV-(dpaea)₂ standards.

Aqueous U^V reached its maximum after 24 hr, after which the concentration decreased steadily over fifty days, concomitant with an increase in solid phase U^IV (Fig. 1a). This suggests that the second electron transfer proceeds much more slowly than the first. Previous work confirms that reduction from U^V to U^IV is indeed mediated by electron transfer from S. oneidensis, as opposed to U^V disproportionation (Molinas et al., 2021

, 2023

). It is likely that reduction of U^VIO₂-dpaea proceeds via dissolution of the solid uranyl^VI followed by rapid reduction of aqueous uranyl^VI, i.e. dissolution is the rate limiting step for the first electron transfer (Molinas et al., 2023

).A slow second electron transfer step (U^V/U^IV) is consistent with abiotic reduction by sodium hydrosulfite (Faizova et al., 2020

Faizova, R., Fadaei‐Tirani, F., Bernier‐Latmani, R., Mazzanti, M. (2020) Ligand‐supported facile conversion of uranyl(VI) into uranium(IV) in organic and aqueous media. Angewandte Chemie 132, 6822–6825. https://doi.org/10.1002/ange.201916334

). Cyclic voltammograms of a U^VO₂-dpaea complex at pH 7 did not display a U^V/U^IV reduction event, suggesting slow electron transfer kinetics that may be related to required structural re-arrangements for the formation of a tri-nuclear U^IV product (Faizova et al., 2018

, 2020

).Uranium isotopic fractionation during the first electron transfer from U^VI to U^V was investigated with a dedicated incubation of U^VIO₂-dpaea (Fig. 2a). Here, the increasing aqueous U showed negligible changes in δ²³⁸U, indicating that the U^VI/U^V reduction displayed little fractionation (Fig. 2b). Reduction of U^VI by a range of bacterial species typically display enrichment of the heavier ²³⁸U in the reduced product, consistent with the predictions of NFS theory during equilibrium isotope fractionation (Basu et al., 2014

). Indeed, ab initio calculation of the expected isotope fractionation factor between the U^VIO₂-dpaea and U^VO₂-dpaea^- at equilibrium gave a value of 0.82–1.60 ‰ (Table S-1), wherein the positive value signals preferential reduction of ²³⁸U. Rather, the isotope signatures of the U^VO₂-dpaea^- observed in the experiment appear consistent with dissolution being the rate limiting step for the first electron transfer, such that U isotope reduction is rapid and quantitative. As dissolution does not involve a redox reaction, the mass-independent isotope fractionation predicted by the NFSE would not be expected.

**Figure 2** **(a)** Aqueous uranium concentrations throughout the first 24 hr of incubation of U^VIO₂-dpaea and *S. oneidensis*. Symbols and error bars depict one standard deviation of the mean of duplicate reactors. **(b)** Corresponding δ²³⁸U values of the aqueous U in duplicate systems (A and B), reported as a fraction of the maximum aqueous U concentration. Symbols and error bars depict two standard deviations of the mean of triplicate measurements. The δ²³⁸U value of the initial U^VIO₂-dpaea is plotted as a yellow dotted line. **(c)** Aqueous uranium concentrations throughout the whole reaction between U^VIO₂-dpaea and *S. oneidensis*. Symbols and error bars depict one standard deviation of the mean of duplicate reactors. **(d)** δ²³⁸U values of the aqueous U after 24 hr when the aqueous U concentration began to decrease. Values are reported as a fraction of the maximum aqueous U concentration. Symbols and error bars depict two standard deviations of the mean of triplicate measurements. The Rayleigh model (blue dashed line) corresponds to the linear best fit of the logarithmic data, R² = 0.89, from which the isotope enrichment factor, ɛ, is derived.

Once U^VI was completely reduced and aqueous U^V reached its maximum concentration after 24 hr, the isotope signature of the aqueous U^V was measured to quantify fractionation during the U^V/U^IV reduction step (Figs. 2d, S-3). Although limited fractionation was observed, Rayleigh distillation models could be fitted to the data, indicating fractionation factors (ɛ) of −0.10 ‰ and −0.11 ‰ for the two batch replicates. These negative values indicate the preferential accumulation of lighter ²³⁵U in the reduced product, contrary to previous observations for microbial U^VI reduction and at odds with NFS theory (Basu et al., 2014

; Stirling et al., 2015

; Stylo et al., 2015

). To ascertain whether this direction of fractionation reflected equilibrium in the peculiar case of a strong aminocarboxylate ligand, we performed ab initio calculations of the fractionation factor at equilibrium between U^VO₂-dpaea⁻ and either U^IV-(dpaea)₂ or a non-uraninite U^IV species, the two likely products of this biological reaction (Molinas et al., 2021

). We modelled the non-uraninite U^IV as a cluster of ningyoite (CaU(PO₄)₂), a close analogue of the non-crystalline biotic reduction products (Bernier-Latmani et al., 2010

Bernier-Latmani, R., Veeramani, H., Vecchia, E.D., Junier, P., Lezama-Pacheco, J.S., Suvorova, E.I., Sharp, J.O., Wigginton, N.S., Bargar, J.R. (2010) Non-uraninite products of microbial U (VI) reduction. Environmental Science and Technology 44, 9456–9462. https://doi.org/10.1021/es101675a

; Alessi et al., 2014

Alessi, D.S., Lezama-Pacheco, J.S., Stubbs, J.E., Janousch, M., Bargar, J.R., Persson, P., Bernier-Latmani, R. (2014) The product of microbial uranium reduction includes multiple species with U(IV)-phosphate coordination. Geochimica et Cosmochimica Acta 131, 115–127. https://doi.org/10.1016/j.gca.2014.01.005

). The fractionation factors of 0.27–0.33 ‰ for the U^IV-(dpaea)₂ product and 0.13–0.46 ‰ for ningyoite both reveal that ²³⁸U would be enriched in the U^IV product at equilibrium (Table S-1), contrary to that observed during biological reduction. These calculations indicate that the bioreduction system was far from equilibrium and suggest that the reaction mechanism precluded the full expression of NFSE that would have enriched ²³⁸U in the product. Furthermore, recent work has proposed that slow microbial reduction should impart significant mass-independent fractionation of up to +1 ‰ (Brown et al., 2018

; Basu et al., 2020

), whereas negative fractionation factors are typically only observed for rapid abiotic reductions, on the order of hours (Stylo et al., 2015

). The slow reduction of the U^VO₂-dpaea^- observed in our experiments (on the order of months), suggests that the proposed reduction rate-fractionation relationship does not hold for all circumstances.To investigate whether equilibrium isotope exchange and the associated expression of the NFSE could overprint the reduction-derived MDF signature, we performed isotope exchange experiments between the solid U^IV product of the bioreduction experiment, with an initial (light) δ²³⁸U of 0 ‰, and aqueous U^VO₂-dpaea^-, with an initial (heavy) δ²³⁸U of ∼5 ‰. Over 200 days, aqueous U became isotopically lighter by 0.6 ‰, indicating the preferential accumulation of ²³⁸U in the U^IV solid (Fig. 3). Whilst this direction of fractionation is in agreement with that calculated for equilibrium, isotope mass balance calculations indicate that the U^IV solid did not become heavier than the aqueous U^V, contrary to the computed equilibrium. These data show that progress to full equilibrium is significantly limited over the course of the experiment, presumably due to slow ligand exchange kinetics.

**Figure 3** **(a)** Aqueous uranium concentrations during equilibrium isotope exchange experiments between U^VO₂-dpaea^– with an initial isotopic composition of ∼5 ‰, and U^IV present as the product of the bioreduction experiments of natural U, with an initial isotopic composition of 0 ‰. Symbols and error bars depict one standard deviation of the mean of duplicate reactors. **(b)** δ²³⁸U values of the aqueous U. Symbols and error bars depict one standard deviation of the mean of duplicate reactors.

This hypothesis is consistent with the strong pentadentate coordination of U^V by dpaea, which provides protection from ligand dissociation and cation-cation interactions typical of U^V disproportionation (Faizova et al., 2018

). Furthermore, any preferential re-oxidation of ²³⁵U^IV to U^V would require the de novo formation of the two uranyl dioxo bonds and re-coordination with dpaea. This is likely kinetically limited due to steric hinderance by the U^IV coordinating ligands. Therefore, we propose that isotope signatures indicating mass-dependent fractionation (faster reaction of ²³⁵U) are preserved during the biological reduction of U^VO₂-dpaea^- because subsequent equilibrium isotope exchange, fractionating in the opposite direction, is limited.Regardless of the abiotic equilibrium isotope exchange between reactants and products (independent of the bioreduction reaction), a recent model has demonstrated the importance of back reaction within the U^VI bioreduction pathway in controlling the overall isotope fractionation (Sato et al., 2021

). The model stipulates that the overall isotope fractionation at each reaction step arises from the balance between the forward and backward reaction rates, and the attendant isotope fractionation for the forward and backward reactions. As such, reactions with equal forward and backward reaction rates will display the full fractionation factor predicted for equilibrium (typically positive for U reduction, indicating preferential accumulation of ²³⁸U in the product). On the other hand, irreversible reactions will result in no observed fractionation. The theory of this model has been demonstrated experimentally during U^VI reduction by S. oneidensis, in which back reaction (reverse electron transfer) was limited by large electron fluxes from oxidation of the electron donor (Brown et al., 2023a

Brown, A.R., Molinas, M., Roebbert, Y., Sato, A., Abe, M., Weyer, S., Bernier-Latmani, R. (2023a) Electron flux is a key determinant of uranium isotope fractionation during bacterial reduction. Communications Earth and Environment 4, 329. https://doi.org/10.1038/s43247-023-00989-x

). These systems result in significantly less isotope fractionation than those with small electron fluxes, which permit more back reaction. The theoretical model and associated experimental evidence, coupled to our observations of the isotope exchange experiment, would suggest that back reaction during biological reduction of U^VO₂-dpaea^- is limited and point toward the role of the U coordinating ligand in controlling the magnitude of isotope fractionation.Furthermore, during microbiological reduction of U^VI-carbonate, the conventional isotopic mass effect was fully expressed, while the NFSE was not (Brown et al., 2023b

Brown, A.R., Roebbert, Y., Sato, A., Hada, M., Abe, M., Weyer, S., Bernier-Latmani, R. (2023b) Contribution of the nuclear field shift to kinetic uranium isotope fractionation. Geochemical Perspectives Letters 27, 43–47. https://doi.org/10.7185/geochemlet.2333

). This implies that the mass-dependent vibrational effect and the mass-independent NFSE are two competing effects operating in opposing directions and is consistent with the proposal that the NFSE requires reaction reversibility in order to overprint the mass-dependent effect.Collectively, these studies indicate that the inhibition of back reaction in the dpaea system is so severe that the mass-dependent isotope fractionation factor is preserved. More specifically, we propose the following mechanism: first, the flux of electrons from the cell to the outer membrane U-reducing proteins is significantly greater than the U^V reduction rates (limited by either low redox potential and/or steric hinderance) (Fig. 4). This allows the redox-active Fe-bearing heme groups of these proteins to become fully reduced prior to electron transfer to U^V. Eventually, electron transfer from the heme Fe^II to U^V occurs with isotopic fractionation according to the conventional mass effect – faster reaction of ²³⁵U. Concurrently, a rapid continuous flux of electrons from metabolism re-reduces the Fe^III of the heme group (empty circles in Fig. 4) and prevents reverse electron transfer from the newly reduced U^IV. Consequently, isotopic equilibration that is dominated by the mass-independent NFSE cannot over-print the initial MDF, unlike in U-carbonate containing systems.

**Figure 4** Cartoon of the proposed mechanism of U isotope reduction and fractionation for both U^VI-carbonate (left) and U^V-dpaea^– (right). Electrons are transferred from the cell to outer membrane U reducing proteins (blue areas) containing multiple redox active heme iron centres (red circles). Depending on the flux of electrons, the heme iron centres are either in their reduced state (solid fill) or oxidised state (open fill).

Likewise, back reaction may also be limited by U^IV sequestration, i.e. kinetic limitations imposed by the U^IV structure and bond rearrangement to recover the uranyl bond structure, resulting in significantly faster electron transfer rates from the heme Fe^II to U^VO₂-dpaea^- than U^IV to heme Fe^III. top

Conclusions

We employed the U^V stabilising ligand, dpaea, to trap aqueous U^V and observed, for the first time, the isotopic signature of U^V throughout the bioreduction of U^VI to U^IV. Whilst the observation of a mass-dependent isotope fractionation factor appears to conflict with previous studies of microbial U reduction, this is likely not an artefact of the unique properties of dpaea (i.e. its ability to solubilise and trap U^V). Rather, these adventitious properties have elucidated the control U coordinating ligands exert over the balance between reactant U supply, electron transfer rate, and U^IV product sequestration. Thus, we infer that other ligands (that cannot stabilise U^V) will behave similarly when such conditions are met. This has significant implications for the interpretation of U isotope signatures in environments where the availability of high affinity ligands may impact U lability. For example, in reducing environments with considerable amounts of organic carbon (providing both a source of electrons for microbial U^VI reduction and a supply of organic complexants), the contribution of the NFSE to observed isotopic signatures may be diminished. This may lead to false interpretations of U isotope signatures, e.g., in applications using organic-rich anoxic sediments as a palaeo-redox archive. In such studies, the observation of lower δ²³⁸U (arising from NFSE-dominated mass-independent fractionation) is usually thought to indicate either a local shift in depositional conditions or water column stratification (Andersen et al., 2017

Andersen, M.B., Stirling, C.H., Weyer, S. (2017) Uranium Isotope Fractionation. Reviews in Mineralogy and Geochemistry 82, 799–850. https://doi.org/10.2138/rmg.2017.82.19

; Brüske et al., 2020

Brüske, A., Weyer, S., Zhao, M.Y., Planavsky, N.J., Wegwerth, A., Neubert, N., Dellwig, O., Lau, K.V., Lyons, T.W. (2020) Correlated molybdenum and uranium isotope signatures in modern anoxic sediments: Implications for their use as paleo-redox proxy. Geochimica et Cosmochimica Acta 270, 449–474. https://doi.org/10.1016/j.gca.2019.11.031

; Lau et al., 2022

Lau, K.V, Hancock, L.G., Severmann, S., Kuzminov, A., Cole, D.B., Behl, R.J., Planavsky, N.J., Lyons, T.W. (2022) Variable local basin hydrography and productivity control the uranium isotope paleoredox proxy in anoxic black shales. Geochimica et Cosmochimica Acta. 317, 433–456. https://doi.org/10.1016/j.gca.2021.10.011

), or a shift in the U isotope mass balance, resulting from enhanced oceanic anoxic environments at regional or global scales (Montoya-Pino et al., 2010

; Andersen et al., 2017

Andersen, M.B., Stirling, C.H., Weyer, S. (2017) Uranium Isotope Fractionation. Reviews in Mineralogy and Geochemistry 82, 799–850. https://doi.org/10.2138/rmg.2017.82.19

). However, our results show that the extent and direction of U isotope fractionation during U reduction may depend on the stabilisation of U^V and, more generally, the lability of U complexes.Furthermore, this study suggests that full expression of isotopic equilibrium in nature may be precluded by U speciation, in addition to the previous roles reported for electron flux and U supply dynamics (Basu et al., 2020

; Brown et al., 2023a

). Future work should focus on delineating these controls with an aim to incorporate U speciation as a parameter within models of U isotope fractionation in the environment. top

Acknowledgements

Funding for this work was provided by an ERC consolidator grant awarded to RB-L (725675: UNEARTH: “Uranium isotope fractionation: a novel biosignature to identify microbial metabolism on early Earth”). This work was also supported by JSPS KAKENHI Grant Numbers JP19K22171, JP21H01864 and JP22J12551. A part of the calculations was performed at the Research Center for Computational Science, Okazaki, Japan (Project: 21-IMS-C049 and 22-IMS-C049). Pierre Rossi provided invaluable support in complying with radiological safety regulations in the lab. We also thank two reviewers and the editor, Claudine Stirling, for their constructive input.Editor: Claudine Stirling top

References

Show in context

We modelled the non-uraninite U^IV as a cluster of ningyoite (CaU(PO₄)₂), a close analogue of the non-crystalline biotic reduction products (Bernier-Latmani et al., 2010; Alessi et al., 2014). View in article

Andersen, M.B., Stirling, C.H., Weyer, S. (2017) Uranium Isotope Fractionation. Reviews in Mineralogy and Geochemistry 82, 799–850. https://doi.org/10.2138/rmg.2017.82.19

Show in context

Such U redox transformations are often accompanied by changes in the ²³⁸U/²³⁵U ratio, reported as δ²³⁸U (Andersen et al., 2017). View in articleAs U isotope fractionation is predominantly associated with redox transformations, U isotope signatures have been utilised as a (1) monitoring tool tuned specifically to the reductive rather than adsorptive removal of U^VI during remediation (Bopp et al., 2010), and (2) palaeo-redox proxy, whereby the preferential reduction of ²³⁸U during marine anoxia is recorded in sedimentary rocks and can be used to reconstruct the pervasiveness of anoxia in past global oceans (Montoya-Pino et al., 2010; Brennecka et al., 2011; Andersen et al., 2017). View in articleIn such studies, the observation of lower δ²³⁸U (arising from NFSE-dominated mass-independent fractionation) is usually thought to indicate either a local shift in depositional conditions or water column stratification (Andersen et al., 2017; Brüske et al., 2020; Lau et al., 2022), or a shift in the U isotope mass balance, resulting from enhanced oceanic anoxic environments at regional or global scales (Montoya-Pino et al., 2010; Andersen et al., 2017). View in article

Show in context

Under anoxic conditions, reduction of U^VI to tetravalent U (U^IV) can be mediated by an array of microorganisms or abiotically via Fe(II)- or sulfide-bearing compounds (Basu et al., 2014; Brown et al., 2018), resulting in the precipitation of sparingly soluble U^IV species. View in articleThus, enrichment of ²³⁸U in U^IV following U^VI reduction has also been attributed to a dominant NFSE (Weyer et al., 2008; Basu et al., 2014, 2020; Stirling et al., 2015; Stylo et al., 2015), despite not necessarily representing isotopic equilibrium conditions. View in articleReduction of U^VI by a range of bacterial species typically display enrichment of the heavier ²³⁸U in the reduced product, consistent with the predictions of NFS theory during equilibrium isotope fractionation (Basu et al., 2014). View in articleThese negative values indicate the preferential accumulation of lighter ²³⁵U in the reduced product, contrary to previous observations for microbial U^VI reduction and at odds with NFS theory (Basu et al., 2014; Stirling et al., 2015; Stylo et al., 2015). View in article

Show in context

Thus, enrichment of ²³⁸U in U^IV following U^VI reduction has also been attributed to a dominant NFSE (Weyer et al., 2008; Basu et al., 2014, 2020; Stirling et al., 2015; Stylo et al., 2015), despite not necessarily representing isotopic equilibrium conditions. View in articleFurthermore, recent work has proposed that slow microbial reduction should impart significant mass-independent fractionation of up to +1 ‰ (Brown et al., 2018; Basu et al., 2020), whereas negative fractionation factors are typically only observed for rapid abiotic reductions, on the order of hours (Stylo et al., 2015). View in articleFurthermore, this study suggests that full expression of isotopic equilibrium in nature may be precluded by U speciation, in addition to the previous roles reported for electron flux and U supply dynamics (Basu et al., 2020; Brown et al., 2023a). View in article

Show in context

This mass-independent fractionation arises from the nuclear field shift effect (NFSE), due to differences in the size and shape of the nuclei of heavy element isotopologues (Bigeleisen, 1996; Schauble, 2007). View in articleAt equilibrium, the NFSE is larger than, and operates in the opposite direction to, the conventional mass-dependent isotope effect, whereby the vibrational zero point energy of the lighter isotope leads to its enrichment in U^IV as mass-dependent fractionation (MDF) (Bigeleisen, 1996; Schauble, 2007; Fujii et al., 2009). View in article

Show in context

The theory of this model has been demonstrated experimentally during U^VI reduction by S. oneidensis, in which back reaction (reverse electron transfer) was limited by large electron fluxes from oxidation of the electron donor (Brown et al., 2023a). View in articleFurthermore, this study suggests that full expression of isotopic equilibrium in nature may be precluded by U speciation, in addition to the previous roles reported for electron flux and U supply dynamics (Basu et al., 2020; Brown et al., 2023a). View in article

Show in context

Furthermore, during microbiological reduction of U^VI-carbonate, the conventional isotopic mass effect was fully expressed, while the NFSE was not (Brown et al., 2023b). View in article

Show in context

In such studies, the observation of lower δ²³⁸U (arising from NFSE-dominated mass-independent fractionation) is usually thought to indicate either a local shift in depositional conditions or water column stratification (Andersen et al., 2017; Brüske et al., 2020; Lau et al., 2022), or a shift in the U isotope mass balance, resulting from enhanced oceanic anoxic environments at regional or global scales (Montoya-Pino et al., 2010; Andersen et al., 2017). View in article

Show in context

The aminocarboxylate ligand dpaea²⁻ (dpaeaH₂ = bis(pyridyl-6-methyl-2-carboxylate)-ethylamine) can be used to precipitate both U^VI and U^IV whilst maintaining U^V as an aqueous complex at circumneutral pH (Faizova et al., 2018). View in articleCyclic voltammograms of a U^VO₂-dpaea complex at pH 7 did not display a U^V/U^IV reduction event, suggesting slow electron transfer kinetics that may be related to required structural re-arrangements for the formation of a tri-nuclear U^IV product (Faizova et al., 2018, 2020). View in articleThis hypothesis is consistent with the strong pentadentate coordination of U^V by dpaea, which provides protection from ligand dissociation and cation-cation interactions typical of U^V disproportionation (Faizova et al., 2018). View in article

Show in context

A slow second electron transfer step (U^V/U^IV) is consistent with abiotic reduction by sodium hydrosulfite (Faizova et al., 2020). View in articleCyclic voltammograms of a U^VO₂-dpaea complex at pH 7 did not display a U^V/U^IV reduction event, suggesting slow electron transfer kinetics that may be related to required structural re-arrangements for the formation of a tri-nuclear U^IV product (Faizova et al., 2018, 2020). View in article

Fujii, T., Moynier, F., Albarède, F. (2009) The nuclear field shift effect in chemical exchange reactions. Chemical Geology 267, 139–156. https://doi.org/10.1016/j.chemgeo.2009.06.015

Show in context

At equilibrium, the NFSE is larger than, and operates in the opposite direction to, the conventional mass-dependent isotope effect, whereby the vibrational zero point energy of the lighter isotope leads to its enrichment in U^IV as mass-dependent fractionation (MDF) (Bigeleisen, 1996; Schauble, 2007; Fujii et al., 2009). View in article

Show in context

During microbiological U^VI reduction, two distinct mechanisms for the complete reduction to U^IV can occur: either via disproportionation of two uranyl^V atoms (generating U^VI and U^IV) (Vettese et al., 2020), or via a second biologically mediated electron transfer to U^V (Molinas et al., 2021, 2023). View in articleThese properties have allowed the reduction of U^VI by Shewanella oneidensis to be followed, revealing the potential for the biological reduction of the U^V intermediate, rather than its disproportionation (Molinas et al., 2021, 2023). View in articleCollectively, these data suggest that the first electron transfer was achieved rapidly, leading to the accumulation of U^V in solution, in agreement with previous studies (Molinas et al., 2021, 2023). View in articlePrevious work confirms that reduction from U^V to U^IV is indeed mediated by electron transfer from S. oneidensis, as opposed to U^V disproportionation (Molinas et al., 2021, 2023). View in articleTo ascertain whether this direction of fractionation reflected equilibrium in the peculiar case of a strong aminocarboxylate ligand, we performed ab initio calculations of the fractionation factor at equilibrium between U^VO₂-dpaea⁻ and either U^IV-(dpaea)₂ or a non-uraninite U^IV species, the two likely products of this biological reaction (Molinas et al., 2021). View in article

Show in context

During microbiological U^VI reduction, two distinct mechanisms for the complete reduction to U^IV can occur: either via disproportionation of two uranyl^V atoms (generating U^VI and U^IV) (Vettese et al., 2020), or via a second biologically mediated electron transfer to U^V (Molinas et al., 2021, 2023). View in articleThese properties have allowed the reduction of U^VI by Shewanella oneidensis to be followed, revealing the potential for the biological reduction of the U^V intermediate, rather than its disproportionation (Molinas et al., 2021, 2023). View in articleCollectively, these data suggest that the first electron transfer was achieved rapidly, leading to the accumulation of U^V in solution, in agreement with previous studies (Molinas et al., 2021, 2023). View in articlePrevious work confirms that reduction from U^V to U^IV is indeed mediated by electron transfer from S. oneidensis, as opposed to U^V disproportionation (Molinas et al., 2021, 2023). View in articleIt is likely that reduction of U^VIO₂-dpaea proceeds via dissolution of the solid uranyl^VI followed by rapid reduction of aqueous uranyl^VI, i.e. dissolution is the rate limiting step for the first electron transfer (Molinas et al., 2023). View in article

Show in context

As U isotope fractionation is predominantly associated with redox transformations, U isotope signatures have been utilised as a (1) monitoring tool tuned specifically to the reductive rather than adsorptive removal of U^VI during remediation (Bopp et al., 2010), and (2) palaeo-redox proxy, whereby the preferential reduction of ²³⁸U during marine anoxia is recorded in sedimentary rocks and can be used to reconstruct the pervasiveness of anoxia in past global oceans (Montoya-Pino et al., 2010; Brennecka et al., 2011; Andersen et al., 2017). View in articleIn such studies, the observation of lower δ²³⁸U (arising from NFSE-dominated mass-independent fractionation) is usually thought to indicate either a local shift in depositional conditions or water column stratification (Andersen et al., 2017; Brüske et al., 2020; Lau et al., 2022), or a shift in the U isotope mass balance, resulting from enhanced oceanic anoxic environments at regional or global scales (Montoya-Pino et al., 2010; Andersen et al., 2017). View in article

Show in context

However, there is increasing evidence of the stabilisation and persistence of U^V intermediates within abiotic and biological systems (Roberts et al., 2017; Pan et al., 2020). View in article

Show in context

However, there is increasing evidence of the stabilisation and persistence of U^V intermediates within abiotic and biological systems (Roberts et al., 2017; Pan et al., 2020). View in article

Show in context

Ab initio calculations of the equilibrium isotope fractionation factor combined with a multi-step model of biological U^VI-carbonate reduction suggests that fractionation factors of up to 1.6 ‰ for the U^VI to U^V step and ∼0.8 ‰ for the U^V to U^IV step (a total of ∼2.4 ‰) may be expected (Sato et al., 2021). View in articleRegardless of the abiotic equilibrium isotope exchange between reactants and products (independent of the bioreduction reaction), a recent model has demonstrated the importance of back reaction within the U^VI bioreduction pathway in controlling the overall isotope fractionation (Sato et al., 2021). View in article

Show in context

Both ab initio calculations and isotope exchange experiments indicate that, at equilibrium, heavy ²³⁸U is enriched in the U^IV oxidation state (Schauble, 2007; Abe et al., 2008; Wang et al., 2015). View in articleThis mass-independent fractionation arises from the nuclear field shift effect (NFSE), due to differences in the size and shape of the nuclei of heavy element isotopologues (Bigeleisen, 1996; Schauble, 2007). View in articleAt equilibrium, the NFSE is larger than, and operates in the opposite direction to, the conventional mass-dependent isotope effect, whereby the vibrational zero point energy of the lighter isotope leads to its enrichment in U^IV as mass-dependent fractionation (MDF) (Bigeleisen, 1996; Schauble, 2007; Fujii et al., 2009). View in article

Show in context

Thus, enrichment of ²³⁸U in U^IV following U^VI reduction has also been attributed to a dominant NFSE (Weyer et al., 2008; Basu et al., 2014, 2020; Stirling et al., 2015; Stylo et al., 2015), despite not necessarily representing isotopic equilibrium conditions. View in articleThese negative values indicate the preferential accumulation of lighter ²³⁵U in the reduced product, contrary to previous observations for microbial U^VI reduction and at odds with NFS theory (Basu et al., 2014; Stirling et al., 2015; Stylo et al., 2015). View in articleFurthermore, recent work has proposed that slow microbial reduction should impart significant mass-independent fractionation of up to +1 ‰ (Brown et al., 2018; Basu et al., 2020), whereas negative fractionation factors are typically only observed for rapid abiotic reductions, on the order of hours (Stylo et al., 2015). View in article