ammonia oxidation, external nitrite must be
present (Fig. 3E).
The hypothesized pathway extends the range
of AOA activity into functionally anoxic en-
vironments. By showing thatN. maritimus
produces both oxygen and N 2 under oxygen
depletion, we have also uncovered a poten-
tially environmentally significant pathway
of N 2 production. By converting ammonium
through nitrite to N 2 , AOA perform a thus far
unrecognized pathway of ammonia conversion
to N 2 when oxygen is depleted.^15 N-tracer ex-
periments currently performed to measure
N-cycling rates in marine oxygen-depleted
environments would overlook this pathway
and account for its N 2 and N 2 O production as
canonical denitrification and/or anammox.
Depending on the environmental controls
and dynamics of this pathway, which are
yet to be elucidated, our current understand-
ing of how oxygen concentrations influence
nitrogen and nitrous oxide production may
require revision.
Considering the abundance of AOA in
the global oceans and their key role in the
nitrogen cycle, a widely distributed oxygen-
producing pathway by AOA could have far-
reaching implications for the microbial ecology
and biogeochemical cycling in oxygen-depleted
environments. The discovery of an oxygen-
producing pathway in AOA can explain the
presence and role of AOA in such environ-
ments, solving a longstanding enigma.REFERENCES AND NOTES- C. Schleper, G. W. Nicol,Adv. Microb. Physiol. 57 ,1– 41
(2010).
2. D. A. Stahl, J. R. de la Torre,Annu. Rev. Microbiol. 66 , 83– 101
(2012).
3. F. J. Stewart, O. Ulloa, E. F. DeLong,Environ. Microbiol. 14 ,
23 – 40 (2012).
4. V. Molina, L. Belmar, O. Ulloa,Environ. Microbiol. 12 ,
2450 – 2465 (2010).
5. M. Sollai, L. Villanueva, E. C. Hopmans, G. J. Reichart,
J. S. Sinninghe Damsté,Geobiology 17 , 91– 109
(2019).
6. C. Berg, V. Vandieken, B. Thamdrup, K. Jürgens,ISME J. 9 ,
1319 – 1332 (2015).
7. J. M. Beman, V. J. Bertics, T. Braunschweiler, J. M. Wilson,
Front. Microbiol. 3 , 263 (2012).
8. P. Lehneret al.,PLOS ONE 10 , e0128125 (2015).
9. L. Tiano, E. Garcia-Robledo, N. P. Revsbech,PLOS ONE 9 ,
e105399 (2014).
10. P. Lam, M. M. M. Kuypers,Annu. Rev. Mar. Sci. 3 , 317– 345
(2011).
11. M. Könnekeet al.,Nature 437 , 543–546 (2005).
12. K. F. Ettwiget al.,Front. Microbiol. 3 , 273 (2012).
13. J.-G. Kimet al.,Proc. Natl. Acad. Sci. U.S.A. 113 , 7888– 7893
(2016).
14. M. T. Wilson, G. Antonini, F. Malatesta, P. Sarti, M. Brunori,
J. Biol. Chem. 269 , 24114–24119 (1994).
15. A. E. Santoroet al.,Global Biogeochem. Cycles 35 ,1– 21
(2021).
16. C. R. Löscheret al.,Biogeosciences 9 , 2419– 2429
(2012).
17. L. A. Bristowet al.,Nat. Geosci. 10 , 24–29 (2017).
18. M. M. Jensen, M. M. M. Kuypers, G. Lavik, B. Thamdrup,
Limnol. Oceanogr. 53 , 23–36 (2008).
19. W. Martens-Habbenaet al.,Environ. Microbiol. 17 , 2261– 2274
(2015).
20. K. F. Ettwiget al.,Nature 464 , 543–548 (2010).
21. C. B. Walkeret al.,Proc. Natl. Acad. Sci. U.S.A. 107 , 8818– 8823
(2010).
22. A. T. Fernandeset al.,FEBS J. 277 , 3176– 3189
(2010).
23. D. M. Arciero, B. S. Pierce, M. P. Hendrich, A. B. Hooper,
Biochemistry 41 , 1703–1709 (2002).
24. M. Stieglmeieret al.,ISME J. 8 , 1135–1146 (2014).
25. Q. Ji, E. Buitenhuis, P. Suntharalingam, J. L. Sarmiento,
B. B. Ward,Global Biogeochem. Cycles 32 , 1790– 1802
(2018).
26. A. E. Santoro, C. Buchwald, M. R. McIlvin, K. L. Casciotti,
Science 333 , 1282–1285 (2011).
27. B. Bayeret al.,Environ. Microbiol. 21 , 4062– 4075
(2019).
ACKNOWLEDGMENTS
We thank A. Glud for assistance with microelectrode measurements
and providing microelectrodes. D.E.C. is a Villum investigator.
Funding:This work was supported by the Villum Foundation,
Denmark (Villum Young Investigator Grant 25491 to B.K.
and Villum Investigator Grant 16518 to D.E.C.); the Independent
Research Fund Denmark (grant 14181-00025 to D.E.C.); and
the Heisenberg Program of the Deutsche Forschungsgemeinschaft
awarded (KO 3651/6-1) to M.K.Author contributions:B.K.
and D.E.C. designed the experiments. B.K. performed the
experiments and analyzed data with input from M.L., L.A.B.,
M.K., B.T., and D.E.C. B.K. and D.E.C. wrote the manuscript,
with contributions and approval from all other authors.
Competing interests:The authors declare no conflicts of
interest.Data and materials availability:All data are available
in the main text or the supplementary materials.SUPPLEMENTARY MATERIALS
science.org/doi/10.1126/science.abe6733
Materials and Methods
Supplementary Text
Figs. S1 to S15
References ( 28 – 49 )
MDAR Reproducibility Checklist21 November 2020; accepted 28 October 2021
10.1126/science.abe6733100 7 JANUARY 2022¥VOL 375 ISSUE 6576 science.orgSCIENCE
Fig. 3. N 2 production and N 2 O accumulation by
N. maritimusduring simultaneous oxygen
production.After sparging with argon,N. maritimus
culture was incubated with either^15 N-nitrite or(^15) N-ammonium and the production of (A) (^30) N 2 and
(B)^29 N 2 was tracked in two independent sets of
incubations for each tracer addition (3 or 4 replicates
each). N 2 production was only detected in incubations
with^15 N-nitrite. For the corresponding oxygen measure-
ments, see figs. S12 and 13. (C) Oxygen accumulation
and^30 N 2 production for a single replicate of the
incubation series 1). The inset shows^30 N 2 production
in the first 20 hours. Disturbances in the oxygen time
series at 0, 3, 10, 15.5, 23, 34, and 38 hours correspond to
the time points when samples for N 2 analysis were
taken, which led to slight oxygen intrusion. Gray dots,
oxygen; black open squares,^30 N 2 .(D)^46 N 2 O accumula-
tion in incubations 1 and 2 supplied with^14 N-ammonium
and 1 mM^15 N-nitrite. Only^46 N 2 O accumulated in
these incubations with a large^15 N-nitrite pool, indicating
that all produced N 2 O originated from nitrite. Error bars
represent the standard deviation of four (incubations
1 and 3) or three (incubations 2 and 4) replicates.
(E) Proposed pathway of oxygen and dinitrogen
production inN. maritimus. Ammonia oxidation pathway
(gray): For the aerobic oxidation of ammonia to nitrite,
oxygen is needed to activate ammonia oxidation by
the ammonia monooxygenase (AMO); two electrons per
oxidized NH 4 +enter the electron transfer chain and,
under aerobic ammonia oxidation, are transferred to the
terminal oxygen reductase, which reduces 0.5O 2.
Proposed oxygen production pathway (black): Nitrite is
reduced to nitric oxide by the NirK nitrite reductase.
Nitric oxide is dismutated to oxygen and nitrous oxide.
The accumulating oxygen is consumed during ammonia
oxidation. Nitrous oxide is reduced to dinitrogen. This
pathway requires eight electrons per produced O 2.
These electrons may partly be supplied by the ammonia
oxidation reaction, which in return would reduce the
oxygen demand by the ammonia oxidation pathway.
RESEARCH | REPORTS