Oceanic Fault Zones Reconstructed

Home , Transform fault

News & views energy regimes than those of aerobic eukar- mitochondria in eukaryotes. In both cases, res- production, has been replaced or retained in yotes. This kind of metabolism has arisen in piratory capacity was acquired by an anaerobic the ciliate’s hydrogenosomes. Evidence indicat- various types of anaerobic eukaryote, and is host cell through the metabolic integration of ing that the ATP transporter identified by Graf associated with the evolution of mitochon- a proteobacterial endosymbiont, and mecha- and colleagues can export ATP to the ciliate host dria into organelles called hydrogenosomes8. nisms can be identified for energy exchange would help to confirm the proposed symbiotic All hydrogenosomes have, to some degree, between symbiont and host cell. Furthermore, interaction. Further discovery and exploration lost mitochondrial pathways for aerobic res- a substantial reduction of the endosymbiont of similarly surprising symbiotic interactions in piration, and generate hydrogen rather than genome is observed, although mitochondrial poorly explored parts of the microbial world is carbon dioxide and water as an end product genomes are typically either much smaller than certainly an exciting prospect for the future. of their energy-generating pathways. Ciliates the endosymbiont genome observed by Graf are exceptionally effective at adapting to oxy- and colleagues, or have been lost completely William H. Lewis and Thijs J. G. Ettema are in gen-depleted environments, and mitochon- (as is the case for several hydrogenosomes10). the Laboratory of Microbiology, Wageningen drion-to-hydrogenosome transitions have Despite these fascinating similarities, there University and Research, NL-6708 WE occurred independently several times in this are also notable differences. Mitochondrial Wageningen, the Netherlands. W.H.L. is also in group of organisms9. endosymbiosis was a much more ancient the Department of Biochemistry, University of Graf and colleagues investigated an anaero- event, enlisting an archaeal host cell rather Cambridge, Cambridge, UK. bic ciliate belonging to the class Plagiopylea, than a modern eukaryotic cell. Mitochondria, e-mails: [email protected]; found in the deepest layers of Lake Zug in even if now reduced from their original form, [email protected] Switzerland. This environment lacks oxygen or even lost from some present-day eukar- 11 and contains relatively high levels of nitrate. yotes , became an integral part of eukary- 1. Graf, J. S. et al. Nature 591, 445–450 (2021). An initial assessment using microscopy otic cells. Genes inherited from the original 2. Eme, L., Spang, A., Lombard, J., Stairs, C. W. & revealed that, unusually, these ciliates have mitochondrial endosymbiont were often re Ettema, T. J. G. Nature Rev. Microbiol. 15, 711–723 (2017). 3. Zaremba-Niedzwiedzka, K. et al. Nature 541, 353–358 a bacterial endosymbiont (belonging to the targeted to the nuclear genome, and some of (2017). class Gammaproteobacteria), rather than an the proteins these genes encoded adopted var- 4. Roger, A. J., Muñoz-Gómez, S. A. & Kamikawa, R. archaeal endosymbiont that produces meth- ious functions throughout the cell. A similar Curr. Biol. 27, R1177–R1192 (2017). 5. Lyons, T. W., Reinhard, C. T. & Planavsky, N. J. Nature 506, ane, which is the more typical type of endo- level of integration is unlikely for the bacterial 307–315 (2014). symbiont present in anaerobic ciliates. endosymbionts of the ciliate investigated. 6. Betts, H. C. et al. Nature Ecol. Evol. 2, 1556–1562 (2018). DNA-sequencing data for lake-water sam- Nevertheless, it would be interesting to 7. Pittis, A. A. & Gabaldón, T. Nature 531, 101–104 (2016). 8. Embley, T. M. Phil. Trans. R. Soc. B 361, 1055–1067 (2006). ples revealed the presence of genes indicating study whether there has been any relocation 9. Embley, T. M. et al. Proc. R. Soc. Lond. B 262, 87–93 (1995). that the ciliate cells have hydrogenosomes. or repurposing of genes between the host 10. Lewis, W. H. et al. Mol. Biol. Evol. 37, 524–539 (2020). Moreover, the sequencing data indicate that and endosymbiont, and to what extent a 11. Karnkowska, A. et al. Curr. Biol. 26, 1274–1284 (2016). the bacterial endosymbionts have an elec- typical mitochondrial function, such as ATP This article was published online on 3 March 2021. tron-transport chain — a collection of protein complexes for respiration that enable energy Geophysics to be produced by a process called oxidative phosphorylation. Graf et al. propose that the electron acceptor in this chain is nitrate, rather than the oxygen used by aerobic organisms. Oceanic fault zones Consistent with this model, the authors report that rates of denitrification (the microbial pro- reconstructed cess that converts nitrate into nitrogen) were higher in lake-water samples in which ciliates Garrett Ito were present than in those from which ciliates had been removed. Tectonic-plate material is generally thought to be neither The genome of the endosymbiont identi- created nor destroyed at plate boundaries called oceanic fied by Graf and colleagues is notably smaller transform faults. An analysis of sea-floor topography suggests than the genomes of most endosymbionts of microbial eukaryotes, containing a mere that this assumption is incorrect. See p.402 310 protein-encoding genes. Among those, the authors identified a gene that encodes a poten- tial transporter protein for ATP, which they At undersea structures called oceanic spread- in a several-kilometre-wide region called the suggest is used to export ATP from the endo- ing centres, two tectonic plates split apart, and transform deformation zone, the crust gen- symbiont into its host, enabling the ciliate to molten rock from volcanic activity solidifies erated at one spreading segment undergoes use the endosymbiont for energy production to produce the crust of the sea floor. These episodes of thinning and then regrowth as it by ‘breathing’ nitrate. This finding represents spreading centres are separated into indi- drifts towards and past the adjacent segment. a unique example of an endosymbiont that has vidual segments that are tens to hundreds of Grevemeyer and colleagues’ work was ena- contributed the capacity for respiration (albeit kilometres long. At the ends of the segments, bled by international collaborations that have using nitrate instead of oxygen as an electron shearing (side-by-side sliding) of the two supported decades of seagoing expeditions acceptor) to a eukaryote that seemingly plates occurs along plate boundaries known to the world’s oceanic spreading centres. The retains organelles of mitochondrial descent as oceanic transform faults. Since their dis- authors analysed sea-floor topography around (hydrogenosomes) — the ancestral versions of covery in the mid-1960s1, these faults have the intersections between spreading-segment which once performed respiratory functions. been considered as sites where plate mater ends and transform faults globally. At these Intriguingly, several parallels can be drawn ial is neither created nor destroyed. But on intersections, young crust produced at one between the cellular partnership discovered page 402, Grevemeyer et al.2 report that this spreading segment (the ‘proximal’ segment) is by Graf and co-workers and the evolution of description is too simplistic. They show that, adjacent to old crust that has been transported

have a widespread role in this new concept Proximal of transform-fault evolution. Moreover, the spreading segment authors make the striking observation that the amount of shoaling between the TDZ and the FZ seems to be independent of the spread- Oceanic Oceanic plate 1 Distal plate 2 ing rate. This finding suggests that the degree spreading Transform Young J-shaped of volcanic reconstruction is as high at cool, segment fault crust ridge magma-starved slow-spreading systems as it is at hot, magma-rich fast-spreading systems. Grevemeyer and colleagues’ discoveries Fracture zone and interpretations are compelling, but also demand further investigation. For example, Cracks and Old Volcanic valleys crust reconstruction future studies should aim to reconcile the Volcanic activity predicted horizontal stretching of the TDZ with the fact that earthquakes along oceanic transform faults are mainly associated with Crust 7 8 Plate motion shearing rather than stretching . Moreover, Rising Lithosphere further modelling work should examine magma whether the results of the authors’ model per- Asthenosphere sist under more-realistic conditions than those considered in their paper. For instance, such work could consider topography supported Figure 1 | Revised understanding of oceanic fault zones. Oceanic spreading centres are areas at which by dynamic stresses9, surface faults and state- two oceanic tectonic plates, consisting of a lithosphere (rigid layer) and crust, split apart. At spreading of-the-art deformation laws derived from rock centres, magma rises from a low-viscosity layer called the asthenosphere, and volcanic activity produces physics. the crust of the sea floor. Spreading centres are separated into segments connected by plate boundaries Regarding volcanic reconstruction, sea- known as oceanic transform faults. In a region called the transform deformation zone (TDZ; blue shaded floor seismic studies are needed to test the region), young crust created at the ‘proximal’ spreading segment is adjacent to old crust that was originally hypothesized crustal thickening and to rec- produced at the ‘distal’ segment. Grevemeyer et al.2 found that the sea floor rises between the old crust in the TDZ and a connected region known as the fracture zone. They propose that volcanic activity at the oncile discrepancies between the evidence proximal segment rapidly reconstructs the TDZ crust adjacent to the fracture zone. They further suggest presented by Grevemeyer and colleagues 10 that J-shaped ridges provide evidence for this process. The authors’ modelling also predicts that the TDZ and that reported in a previous study — undergoes both shearing (side-by-side sliding; blue arrows) and stretching. The stretching results in cracks especially, evidence from gravity measure- and valleys, and explains why the TDZ is deep. ments that suggests crustal thickening does not occur at slow-spreading systems. Finally, from its place of origin at the other segment mantle)4, the authors determined that the TDZ it is unclear whether the magma that forms (the ‘distal’ segment). also undergoes horizontal stretching. This the J-shaped ridges originates beneath the The authors found that the sea floor of the old stretching thins the lithosphere, causing the spreading centre and propagates laterally, as crust in this transform deformation zone (TDZ) TDZ to deepen. The computer models predict proposed, or is generated locally below the is consistently deeper than the sea floor in the that the amount of stretching and deepening TDZ–FZ intersection, rises mostly vertically fracture zone (FZ) — a scar of fractured crust increases with increasing difference in the and is guided into the J shape by lithospheric where this old crust has already drifted past crustal age across the TDZ. This difference stresses. Resolving this issue will require some the proximal segment, and shearing has ceased increases with decreasing spreading rate combination of modelling, detailed sea-floor (Fig. 1). This rapid shoaling (rise in sea-floor because, at lower rates, the crust from the observations, and sampling and chemical elevation) is opposite to the gradual sinking distal segment takes longer to reach the prox- analysis of the lava rock of the ridges. Such expected, considering that, in migrating from imal segment. Therefore, the authors’ models work is also crucial for an improved under- the TDZ to the FZ, the crust continues to age, explain the newly discovered global relation- standing of the physics of magma generation and should therefore cool and become denser. ship between spreading rate and TDZ depth. and transport near all types of plate boundary. Grevemeyer et al. then examined systems of Grevemeyer et al. attribute the rapid shoal- spreading segments that represent the global ing between the TDZ and the FZ to magma Garrett Ito is in the Department of Earth range of sea-floor-spreading rates. They dis- produced beneath the spreading centre that Sciences, University of Hawaii, Honolulu, covered that the depth of the TDZ tends to be periodically overshoots the proximal-segment Hawaii 96822, USA. maximal when sea-floor spreading is ultraslow3 end (for example, in magma-filled fractures) e-mail: [email protected] (slower than 20 millimetres per year) or slow and reconstructs the TDZ crust. The evidence (20–55 mm yr−1), and minimal when spreading for this overshooting is seen in a series of 1. Wilson, J. T. Nature 207, 343–347 (1965). −1 2. Grevemeyer, I., Rupke, L. H., Morgan, J. P., Iyer, K. & is relatively fast (55–140 mm yr ). To explore J-shaped ridges that run parallel to the prox- Devey, C. W. Nature 591, 402–407 (2021). the physical causes of this relationship, the imal segment and extend across the TDZ–FZ 3. Dick, H. J. B., Lin, J. & Schouten, H. Nature 426, 405–412 authors produced computer models that sim- intersection, hooking inwards towards the TDZ (2003). 4. Behn, M. D., Boettcher, M. S. & Hirth, G. Geology 35, ulate deformation in the volume of crust and (Fig. 1). Further signs of such volcanic recon- 307–310 (2007). mantle below a transform fault joining two struction include circular volcanic domes and 5. Lonsdale, P. Mar. Geophys. Res. 8, 203–242 (1986). spreading segments. other hilly topographic features. 6. Fornari, D. J. et al. Mar. Geophys. Res. 11, 263–299 (1989). 7. Sykes, L. R. J. Geophys. Res. 72, 2131–2153 (1967). As expected, Grevemeyer and colleagues Such J-shaped ridges have been noted in 8. Wolfe, C. J., Bergman, E. A. & Solomon, S. C. J. Geophys. found that the geometry of the plate bound- studies of systems in which the spreading rate Res. Solid Earth 98, 16187–16211 (1993). ary leads to shearing in the TDZ. However, is intermediate5 or fast6. However, Grevemeyer 9. Bercovici, D., Dick, H. J. B. & Wagner, T. P. J. Geophys. Res. Solid Earth 97, 14195–14206 (1992). by accurately simulating deformation of the et al. document the prevalence of these features 10. Gregg, P. M., Lin, J., Behn, M. D. & Montési, L. G. J. Nature brittle lithosphere (the crust and uppermost at all spreading rates, and assert that the ridges 448, 183–187 (2007).