Extremophiles in Earth's Deep Seas: A View Toward Life in Exo-Oceans

2.1. Low and unique energy pathways

One of the most prominent characteristics of the deep oceans is both the low overall chemical energy available to support life and the unusual and diverse pathways in which life acquires carbon energy (McClain et al., 2012a, 2020). Photosynthetically active radiation is virtually nonexistent, and consequently, primary production is virtually absent, occurring only through alternative pathways, such as chemosynthesis. The chemical energy that sustains most deep-sea organisms is sequestered from sinking particulate organic carbon (POC) derived from primary production in the euphotic zone hundreds of meters to kilometers above. POC flux decreases with depth in the water column and distance seaward from productive coastal regions; material is remineralized as it moves from its source (Mouw et al., 2016). At the abyssal seafloor, this downward flux represents less than 1% of surface production (Lampitt and Anita, 1997). POC delivery to the seafloor greater than 2000 m can vary two orders of magnitude from 0.014 to 5.24 gC m⁻² y⁻¹ within deep-sea regions globally (Lampitt and Anita, 1997). The estimates of deep-sea POC flux are considerably lower than the typical flux values for shallower marine systems (<100 m) where POC estimates are typically well over 30 gC m⁻² y⁻¹ (Lutz et al., 2007) and can reach values well over 300 gC m⁻² y⁻¹ (Hung et al., 2013).

The area with the lowest energy available for life may well be the South Pacific Bare Zone, a large region, approximately the size of the Mediterranean Sea, devoid of sediment with little to no delivery of carbon (Rea et al., 2006). The Bare Zone lies beneath the extremely low-productivity South Pacific Gyre, dating to the end of the Cretaceous, and by consequence the rain of POC to the abyssal floor is minimal (Rea et al., 2006). The oceanic isolation of the Bare Zone results in essentially no dust or other terrigenous input of carbon (Rea et al., 2006). The Bare Zone is also far from geologically active areas meaning that carbon input from hydrothermal vents or other chemosynthetic systems is also absent (Rea et al., 2006). Whereas one or two of these conditions are common, the combination of all three is unique to the Bare Zone. Previous research in the region did find microbial communities of the subseafloor but—as predicted—in very low biomass and very low metabolic activity (D'Hondt et al., 2009), with cell abundances three to four orders of magnitude lower than other subseafloor communities (D'Hondt et al., 2009).

Special attention should be directed as well to subsurface deep-sea sediments, those environments below where sediment is reworked due to biological activity (bioturbation zone) (Jones et al., 2018). In general, energy availability in these habitats mirrors the energy availability of seafloor sediments, which is lowest in the oligotrophic regions and decreases as the water column depth increases. Likewise, as sediment depth increases, energy availability also decreases as carbon is mineralized and buried. These subsurface systems appear to be carbon-limited given the preponderance of energy-rich electron acceptors throughout the sediment column (D'Hondt et al., 2004). Yet, prokaryotic biomass in these deep subsurface sediments is approximately equal to that of the ocean's water column (Kallmeyer et al., 2012). Microbial processes are also detectable in these low-energy settings and comparable with those other deep-sea sediments (Orcutt et al., 2013; Jones et al., 2018).

However, the amount of energy while allowing for microbial presence may lead to fundamental shifts in the microbial communities' physiological states (Price and Sowers, 2004; Jones et al., 2018). At the highest energy, growth requires energy beyond basic metabolic demand to allow for cell division and biomolecule synthesis. If energy levels decline, then only basal maintenance can be supported allowing for cells to repair, replace, and maintain biomolecules. One particular limitation in the deep subsurface is amino acid racemization and DNA depurination, which are thought to constrain microbial activities at depth where temperatures increase with a thermal gradient and energetic substrates are lacking (Lomstein et al., 2012). Ultimately, at the lowest energy states, cells must switch to dormancy, a state of low to zero metabolic activity. For some bacterial species, this dormant state results in the formation of endospores, which outnumber vegetative cells in the deep subsurface (Wörmer et al., 2019). Of course, the dormancy strategy is bet hedging on a future scenario where energy availability rises as metabolic maintenance is required to repair cellular damage from the environment (Hoehler and Jørgensen, 2013). The energy thresholds for these physiological states are still being actively researched but likely vary by orders of magnitude [10⁶:10³:1 (Hoehler and Jørgensen, 2013; Jones et al., 2018)].

In general, deep-sea organisms exhibit a broad suite of adaptations to lower total metabolic costs to both persist and thrive in low-energy conditions (McClain et al., 2020), and a number of trait trade-offs are known to covary with food availability in the deep sea. Previous studies have demonstrated, in line with the energy budget theory, that biomass, diversity, and ecosystem processes in deep-sea benthos are all highly correlated with POC flux (Smith et al., 2008; McClain et al., 2012a). With decreasing POC flux and in low-energy environments, there is selection for reduced energetic costs (Hoehler and Jørgensen, 2013). One of the most recognizable ways this occurs is through increasing dominance of prokaryotes and overall smaller eukaryotes with decreasing energy (Thiel, 1975; Rex et al., 2006). In general, deep-sea faunas are miniaturized versions of their shallow water counterparts, with an overall reduction in body size with increased depth and lower food availability (Rex et al., 2006; McClain et al., 2009), although patterns may be more complex within clades (McClain et al., 2012b). For instance, gastropod genera with small-bodied, shallow-water species have significantly larger deep-sea representatives, while the opposite is true for genera that are large bodied in shallow water (McClain et al., 2006), resulting in the loss of both the largest and smallest species with increased depth and lower food availability (McClain et al., 2012b). Larger species are likely lost because a threshold in food availability occurs where both large body sizes and self-sustaining populations cannot be maintained (McClain et al., 2006, 2012b). Constraints on smaller sizes with decreased food availability may represent selection against a reduced foraging area and lower starvation (McClain et al., 2006). In contrast, while bivalves also exhibit constraints against larger sizes at low food availability, no pattern is observed among the smallest sizes, possibly reflecting reduced motility and energetic demands (McClain et al., 2012b). Overall, the body size may be an adaptation to carbon availability, with the relationship mediated by other axes of the metabolic niche such as motility or trophic level.

The overall winnowing or reduction of range in the adaptive landscape from energy limitation appears to be echoed in other traits as well. McClain et al. (2004, 2005) found that with increased depth, and presumed lower energy availability, considerable changes were noticeable within gastropod shells. These shifts included a loss of shell sculpture, for example, spines and ribs, as well as increased frequency of more spherical shells as opposed to more cylindrical shells with increased depth. The production of shell material is metabolically expensive, and at depth where calcium carbonate may dissolve, maintenance of the shell becomes even more costly. Thus, at great abyssal depths where energy availability is low, the maintenance of complex shells is not permissible. This exclusion of metabolically expensive strategies at greater depth is also replicated by the relative absence of nudibranchs in deep-sea settings (McClain, pers. comm.). Nudibranchs in general possess higher metabolic rates and more energetically costly strategies, for example, hermaphroditism, compared with other mollusc groups (McClain et al., 2014). When nudibranchs do occur in larger numbers in the deep sea, they occupy canyon, ridge, and seamount habitats where energy is abundant. These basic examples suggest that low-energy availability may limit the complexity and diversity of life.

Yet, anecdotally the energy limitation of the deep sea appears to have generated evolutionary novelty as well. Deep-sea organisms exhibit a variety of creative solutions to deal with lower food availability. Contrary to the overall winnowing of sizes and one of the best examples in defiance of the general trend of miniaturization is the giant isopod Bathynomus giganteus at 36 cm in length (McClain et al., 2015). It is the largest in the order Isopoda and one of the largest known crustaceans. The larger size of the giant isopod may be an adaptation that allows it to quickly and efficiently monopolize food in the deep sea (McClain et al., 2015). A larger size also confers a greater foraging area to an organism (McClain et al., 2015). The giant isopod can survive eight weeks between feedings (McClain et al., 2015). In aquaria, the gastropod Neptunea amianta, orders of magnitude larger than other deep-sea gastropods, can survive three months between meals (Tamburri and Barry, 1999). This fasting potential reflects the ability of larger organisms to hold greater lipid reserves.

Females in the fish family Ceratiidae may also ensure their survival in a food-limited environment with the organ responsible for the group's common name, angler fish (Pietsch, 2009). Evolutionarily derived from the spines of the dorsal fin, a lighted blue lure rests on a stalk above the forehead (Pietsch, 2009). Symbiotic bacteria at the lure's tip produce light that attracts prey. Siphonophore jellyfish in the genus Erenna also utilizes light to attract prey (Haddock et al., 2005). Red light, a rarity among luminescent organisms, is emitted when the jellyfish flicks its prey-shaped tentacles, making them closely resemble copepods, a food source for many small fish (Haddock et al., 2005).

Unexpected carnivory in typically filter-feeding groups of sponges (Vacelet, 2007) and tunicates (Okuyama et al., 2002) also may indicate evolutionary novelty in response to low food availability. In the group of Cladorhizdae, species of carnivorous sponge, exterior spicules have evolved into hooks that can ensnare small crustaceans (Vacelet, 2007).

Arguably, the lower food availability and resource competition in the deep sea may have driven the evolution of species to exploit and specialize on new carbon resources (Sanders, 1977). This has been documented in deep-sea polychaetes (Jumars et al., 1990) anecdotally in the occurrence of wood- and bone-eating deep-sea specialists (Turner, 1973, 1977; Rouse et al., 2018), as well as the utilization of nonphotosynthetic energy pathways.

In the deep oceans, even photosynthesis has evolved away from requiring light energy, thus indicating that photosynthesis is not limited to solar environments (Beatty et al., 2005). The green sulfur bacteria, GSB1, isolated from hydrothermal vent plume fluids on the East Pacific Rise, possess an in vivo absorption spectrum similar to other green sulfur bacteria. This absorption spectrum, electron microscopy, and DNA sequencing of GSB1 indicate the presence of bacteriochlorophyll c. The occurrence of GSB1 at vents and the presence of bacteriochlorophyll c point to the geothermal radiation as the photon donor to photosynthesis in these bacteria. Indeed, geothermal photon flux at the vent was estimated to be in the same range as solar photo availability to green sulfur bacterium living at 80 m depth in the Black Sea.

In contrast, the discovery of hydrothermal vents and methane seeps has demonstrated that life also occurs without photosynthesis (Chyba and Hand, 2001). Deep-sea metabolism can also be based on methane, sulfur, and iron. Chemolithoautotrophy is prevalent in the deep oceans where organisms utilize inorganic chemical energy sources to obtain metabolic energy, which they in turn use to synthesize biomass from CO₂ and other inorganic compounds. Vent and seep settings jettison reduced chemicals (e.g., hydrogen, sulfide, and methane) into an oxic water column. The thermodynamic gradients created by the simultaneous presence of reduced and oxidized chemicals are one that microbes exploit to create energy (McCollom and Shock, 1997). McNichol et al. (2018) estimated that chemosynthetic productivity was orders of magnitude higher than primary productivity reaching the seafloor at the site. Through a global extrapolation, chemosynthetic production may be over 1.4 Tg C y⁻¹ (McNichol et al., 2018).

2.2. High temperature

Hydrothermal vents are home to the greatest recorded temperatures on Earth. At oceanic ridges, where rocks are often brittle and fractured, cold seawater percolates down through Earth's crust, is superheated by magma, and rises back to the surface (Van Dover, 2000). To date, the “Two Boats” vent in the Turtle Pits field along the Mid-Atlantic Ridge holds the record for the hottest hydrothermal vent. Here, vent fluid temperatures reach up to 464°C (Koschinsky et al., 2008), well above the critical point of sea water (407°C). The second hottest vent on record, “Sisters Peak,” in the Comfortless Cove field, spews vent fluids as hot as 400°C (Haase et al., 2007).

Hyperthermophilic methanogenic archaea, known from deep-sea hydrothermal vents, are potentially the oldest members of archaea on Earth, and represent the majority of hyperthermophiles in the deep sea (Van Dover, 2000; Takai et al., 2008). A species of this group, Methanopyrus kandleri strain 116, from the Kairei hydrothermal field in the Central Indian Ridge, holds the current record for the highest temperature tolerance of any organism. This species grows and divides in waters up to 122°C and survives a brief exposure (3 h) to 130°C water (Takai et al., 2008). Archaea strain 121, a coccoid archaea species collected from the Finn vent within the Mothra vent field has the second highest temperature tolerance (Kashefi and Lovley, 2003), thriving in temperatures up to 121°C and persisting and even remaining reproductively viable after a short exposure time (2 h) at 130°C.

A variety of adaptations allow for the temperature tolerance of these archaea. Incorporation of long-chain lipids into the cell membrane provides significant thermostability and maintenance of membrane integrity (Walsby, 1983; Van Dover, 2000). High-density layers of long-chain lipids have also been detected in pooled 350°C vent fluids (350°C), suggesting the presence of archaea nearby; however, the specific upper temperature limit is yet to be determined (Hedrick et al., 1992). Elevated incorporation of straight-chain and saturated fatty acids into cell membranes allows for the maintenance of proper membrane fluidity under extreme temperatures (Rampelotto, 2010). Heat shock proteins, found in all cells exposed to extremely high temperatures, can destroy thermally degraded or correct thermally induced denaturations in proteins, allowing cellular functions to continue under short exposure to extremely high temperatures (Van Dover, 2000; Rampelotto, 2010). Electrostatic interactions within proteins also aid in stability under high temperatures, and many hyperthermophile proteins possess elevated levels of these interactions (Kumar and Nussinov, 2001). Other modifications to proteins, such as the reduction of mobile surface loops and general size, also aid in the ability to cope with extreme temperatures, as smaller proteins have a smaller heat capacity change resulting in a higher melting temperature (Kumar and Nussinov, 2001; Rampelotto, 2010).

The extreme temperatures of hydrothermal vent sites are also home to metazoans, particularly polychaete worms, with a variety of adaptations to heat. Alvinella pompejana was proposed to be the most eurythermal species on the planet (Cary et al., 1998), and capable of surviving brief exposures to 100°C waters (Chevaldonné et al., 2000). Although suggested that A. pompejana may consistently live in conditions greater than 60°C (Cary et al., 1998), steep thermal gradients and the difficulty of sampling fragile worm tubes from a submersible challenge the accuracy of these temperature measurements (Chevaldonné et al., 2000). Temperatures adjacent to the worm's tubes more likely range between 2°C and 25°C and rarely go over 45°C because of rapid mixing of hot vent fluids and background cold seawater (Chevaldonné et al., 2000). The worm's macromolecules and organelles are thermally stable at temperatures around 50°C also suggesting that the maximum physiological body temperature is below 50°C and extended periods above this temperature would be lethal (Chevaldonné et al., 2000). This is in contrast to statements that functioning could occur above 80°C (Cary et al., 1998). Using high-pressure aquaria and mimicking natural thermal gradients, newer research has quantified the thermal tolerance of both A. pompejana and the related species Paralvinella sulfincola. Experimental work on A. pompejana demonstrated that prolonged exposure between 50°C and 55°C was lethal with the highest survival rates seen at 42°C (Ravaux et al., 2013). P. sulfincola prefers temperatures between 40°C and 50°C and can survive 15 min of acute exposure at 55°C (Girguis and Lee, 2006). This is supported by previous work demonstrating that the thermal limit of P. sulfincola is between 50°C and 56°C (Lee, 2003). Thus, the current evidence places A. pompejana and P. sulfincola tying or exceeding the range of other high-temperature-adapted metazoans (Girguis and Lee, 2006).

This thermotolerance in the Alvinellid worms results from a host of adaptations. For example, cuticle and interstitial collagen of A. pompejana are the most thermally stable natural proteins known to exist, largely due to the fibrillar structure of the proteins in tissues (Gaill, 1993). In addition, the ribosomal DNA of Alvinellid species contains a greater ratio of triple hydrogen-bonded guanine–cytosine pairs to double hydrogen-bonded adenine–thymine pairs than most vent dwelling species, allowing ribosomal DNA to remain stable in temperatures up to 88°C (Dixon et al., 1992). Proteomic investigation into P. sulfincola found the ability to mitigate oxidative stress by increasing the synthesis of antioxidants and decreasing flux through the mitochondrial electron transport chain suggesting that oxidative stress may limit metazoan thermotolerance (Dilly et al., 2012).

2.3. Low temperature

With 70% of Earth's surface covered by oceans with a mean temperature of −1°C to 5°C, most of our planet is consistently cold, making low temperature the most widespread extreme environment on the planet (Casanueva et al., 2010; Siddiqui et al., 2013; De Maayer et al., 2014). As a result, psychrophiles have become the most widely distributed, abundant, and diverse extremophiles on Earth (De Maayer et al., 2014). In the world's oceans, −2°C is the lower temperature boundary, as this is approximately the freezing point of seawater with normal salinity (Clarke et al., 2009). The Southern Ocean, and specifically the Weddell Sea and Ross Sea, is the coldest marine environment on the Earth (Clarke et al., 2009), with seafloor temperatures throughout the year at approximately −2°C. Within the Weddell Sea, the Filchner Trench funnels cold surface waters cooled by surrounding ice shelves, creating the coldest deep water on record with potential temperatures as low as −2.4°C (Ekau, 1990).

No microbial studies within the deep waters of the Filchner Trench have been conducted to date, but microbial life is almost certain despite the lack of data. Prokaryotic communities from nearby deep waters (>2000 m) comprised α-, δ-, and γ-proteobacteria, as well as cytophagales, planctomycetales, and members of Marinimicrobia (López-García et al., 2001). Within the Filchner Trench, invertebrate communities are dominated by polychaetes, bivalves, holothurians, anthozoans, cumaceans, and ophiuroids, with representatives of most other major taxa at lower abundances (Pineda-Metz et al., 2019). Several fish species have been recorded from within the Filchner Trench as well, including Gerlachea australis (the dominant species), Akarotaxis nudiceps (an abundant species), Chaenodraco wilsoni, Pleuragramma antarcticum, and the rarely seen Trematomus tokarevi (La Mesa et al., 2019). Each of these fishes belongs to the Perciform suborder Notothenioidei, whose members compose ∼90% of the total fish abundance and biomass in the Southern Ocean (Ekau, 1990).

At low temperatures, cellular membrane fluidity is greatly reduced, becoming rigid and impacting transport of cellular materials across the membrane. Cold-adapted bacteria and archaea both display elevated levels of unsaturated fatty acids in the cell membrane aiding in membrane fluidity as the double bond in unsaturated fatty acids does not allow for dense packing (Myka et al., 2017). Other consequences of low temperature include reduced diffusion rates, membrane conformation, maintenance of structure, and proper functioning of proteins, and reduced enzymatic activity (Myka et al., 2017). Gamma-proteobacteria (Colwellia spp.) increase the production of extracellular polysaccharide substances with cryoprotectant properties when in sub 0°C waters; likely a major contributor to the group's success in cold environments (Marx et al., 2009). Several deep-sea Psychrobacter species synthesize cold-adapted lipases and esterases that maintain a relatively high activity at temperatures near 0°C (e.g., Wu et al., 2013). Proteins of cold-adapted archaea also contain less hydrophobic polar amino acids (especially Leucine), and higher levels of noncharged polar amino acids allowing for a reduction of activation energy, greater flexibility, and increased catalytic efficiency of proteins at low temperatures (Saunders et al., 2003). Both bacterial and archaeal species isolated from the deep sea produce cold shock proteins when in low-temperature environments, which allow transcription and protein production to continue at cold temperatures (Cavicchioli et al., 2000; Giaquinto et al., 2007).

Polar marine invertebrates generally display reduced basal metabolic rates, energetic demands, growth rates, protein synthesis rates, and reproductive effort when compared with similar species of temperate waters (Clarke, 1980). Polar benthic invertebrates generally avoid pelagic larvae in favor of brooding large-yolked eggs and protected development, which is thought to be an adaptation to both the cold temperatures and low energy of polar regions (Thorson, 1950). Cold-adapted invertebrates also evolved reduced activation energy requirements for enzymatic activity in response to low temperature and low kinetic energy within the cell (Hazel and Prosser, 1974; Clarke, 1980). One of the key adaptations allowing many crustacean species to persist in these cold waters is the ability to downregulate the concentration of Mg²⁺ in the hemolymph; those that do not have this adaptation do not occur in cold habitats due to poor cardiac and ventilatory performance (Thatje et al., 2005).

Antarctic nototheniid fish exhibit multiple adaptations to low temperature explaining, in part, their cold-water dominance. To combat elevated levels of reactive oxygen species, these fish display increased transcription rates of proteins with antioxidant properties (Chen et al., 2008). Low temperatures would increase the viscosity of blood with obvious negative cardiac effects. By decreasing both erythrocytes and hemoglobin, the amount of energy needed for circulation is decreased. For example, nototheniids in the family Channichthyidae have lost hemoglobin production all together, reducing the blood-oxygen carrying capacity by 90%, and counterbalancing the effects of increased heart rate and blood volume due to increasing blood viscosity (Giordano et al., 2015). Other hemoglobin modifications include those of P. antarcticum with three kinds of hemoglobin, each displaying varying thermodynamic behaviors (Di Prisco, 1998). High concentrations of NaCl in the blood of the nototheniid Pagothenia borchgrevinki allow for a lower blood freezing point (−2.7°C) compared with that of tropical fish (−0.8°C) (DeVries, 1988). Antarctic nototheniid fish also contain eight forms of antifreeze glycopeptides and peptides, which bind ice nuclei and depress the freezing point (Ahlgren et al., 1988; Cheng and DeVries, 1991). Blood, pericardial fluid, peritoneal fluid, epidural fluid, and the intestines of these fishes all contain antifreeze peptides. This production of diverse antifreeze forms is thought to be the key adaptation that allowed the nototheniids to diversify and dominate the Southern Ocean (Cheng and DeVries, 1991; Johnston, 2003).

2.4. Hypoxia and anoxia

Like many extreme conditions in the deep sea, limited oxygen is pervasive, occurring in habitats at many scales. Large swaths of oxygen-limited water (Oxygen minimum zones, OMZ) occur between a 10 and 1300 m depth, depending on local circulation patterns, so they often but not always are considered deep-water phenomena, particularly affecting the bathyal zone (Levin, 2003). Oxygen is depleted here by high oxygen consumption by heterotrophic life relative to low oxygen supply via primary production. Within OMZs, oxygen concentrations are less than 0.5 mL l⁻¹ or 22 μM, which corresponds to <7.5% saturation (Levin, 2003). The largest OMZs occur in the East Pacific, the Arabian Sea, the Bay of Bengal, and off the coast of Southwest Africa (Levin, 2003). OMZ phenomena are reviewed in great detail by Levin (2003). Other pockets of the deepest parts of the ocean also experience permanent hypoxia. These include off the coast of Baja, California, the Saanich Inlet, some fjords, deep hypersaline anoxic basins (DHABs), the Santa Barbara Basin in California, and the interior of the Black Sea (Levin, 2003).

Yet another globally distributed, oxygen-depleted environment is the ocean's seafloor sediments; almost all become anoxic at some depth below their surface, generally in the shallowest layers (Diaz and Rosenberg, 1995; Glud, 2008). While the macrofauna exists only in the shallowest layers where bioturbation activity allows some oxygen to penetrate (Glud, 2008), microbial life has been found as deep as 2.5 km below the surface of marine sediments (Inagaki et al., 2015), and global estimates of microbial biomass in these environments are upward of 90 Pg C (Lipp et al., 2008). Archaea dominate the microbial biomass of these environments relative to bacteria, likely due to the lower energy requirements of archaeal cells relative to bacterial cells (Lipp et al., 2008).

Oxygen is central to life on Earth because it is an excellent electron acceptor in cellular respiration. However, cellular respiration can occur in the absence of oxygen. Many prokaryotes have developed the ability to undergo anaerobic cellular respiration, either facultatively or as an obligate anaerobe. These prokaryotes commonly use sulfate, nitrate, manganese oxides, iron oxides, or methane as an electron acceptor (Glud, 2008). Furthermore, evidence of the first metazoans, three species of Loriciferans, known to be capable of persisting in permanent anoxic conditions, comes from DHABs (Danovaro et al., 2010). While these Loriciferans can thrive without oxygen, with regard to energy metabolism, all metazoans need oxygen for cholesterol synthesis (Brown and Galea, 2010). Further research is needed to understand how these species circumvent this requirement, for example, recovery of cholesterol through diet. Eukaryotic life may have originated as symbiosis between methanogenic archaea and B-proteobacteria in anaerobic environments (Moreira and López-García, 1998) and these environments still host an abundance of symbioses, leading to the hypotheses that deep oxygen-depleted environments are excellent candidate sites for eukaryotic diversification (Bernhard et al., 2000).

Prokaryotic species in OMZs have adapted to low oxygen concentrations by eliminating the need for oxygen in cellular functions altogether. Instead, prokaryotic anaerobes utilize nitrate, sulfur, magnesium oxide, iron oxide, ammonia, or methane as oxidants for metabolism (Glud, 2008; Berg et al., 2015). Foraminifera in dysaerobic environments tend to have smaller and thinner tests with larger and more numerous test pores than those found in well-oxygenated waters (Perez-Cruz and Machain-Castillo, 1990; Gooday et al., 2000; Gibson and Atkinson, 2003). This decreases oxygen requirements and increases the surface area:volume ratio for more efficient respiration (Gibson and Atkinson, 2003). Additional and larger pores in the tests may allow for a greater absorption of oxygen from the water column, as mitochondria of foraminiferans from low-oxygen environments aggregate around pore openings and in the apertural cytoplasm, which forms the extendable pseudopods (Perez-Cruz and Machain-Castillo, 1990; Bernhard and Gupta, 1999).

Among eukaryotes, adaptations to low oxygen availability fall categorically into strategies that lower the need for or increase availability thereof. Oxygen availability has clearly been the lynch pin in the progression of animal evolution (Payne et al., 2008; Sperling et al., 2015). Dramatic increases in body size occur over the 3.5-million-year evolution of life when atmospheric oxygen increases (Payne et al., 2008, 2010). These long-temporal linkages between size and oxygen are also mirrored over spatial gradients in the deep sea, with organisms typically being larger in more oxygenated regions (McClain and Rex, 2001). Certain ecological traits, such as carnivory, may also be tied to higher oxygen levels as active predation and digestion of large prey parcels are metabolically expensive (Sperling et al., 2013). It is worth noting that many of those traits favored in low chemical energy environments would also be favored in low-oxygen environments as well, as both reduce metabolic demand.

Among eukaryotes, adaptations to increase oxygen availability are equally diverse. Nematodes and annelids found in high densities in various OMZs possess elongated body plans providing high surface area-to-volume ratios (Levin, 2002). In addition, annelids from deep-sea hypoxic environments have been found to possess multiple types of hemoglobins with much greater oxygen affinities than those of well-oxygenated waters (Hourdez and Weber, 2005). Crustaceans inhabiting chronic hypoxic waters near hydrothermal vents possess enhanced gill surface areas achieved through more numerous lamellae (Decelle et al., 2010). The crustacean, Gnathophausia ingens, in addition to increased gill surface area, possesses a highly efficient hemocyanin respiratory pigment that can remove up to 90% of dissolved oxygen from water passing over the gills (Sanders and Childress, 1990). Upregulation of hypoxia-inducing factor-1α, regulating genes responsible for breathing, vasodilation, anaerobic metabolism, and the development of new red blood cells and blood vessels, is the main response to hypoxia in fishes (Zhu et al., 2013). Glucose transporter proteins are also upregulated when in hypoxic waters and induce cellular glucose uptake, thought to be a response to satisfy increased energy demands of the gills when in hypoxic conditions (Zhu et al., 2013).

3.1. High pressure

The highest known pressure in the deep oceans coincides with the maximum known depth in the southern end of the Mariana Trench, the Challenger Deep. The actual maximum depth of Challenger Deep is disputed. The deepest reported measurement was by a Russian research vessel at 11,034 m (Taira et al., 2004). The International Hydrographic Organization adopted 10,924 in 1993 (Taira et al., 2004), but a subsequent 2010 survey, and the most accurate, places the depth at 10,984 ± 25 m (Gardner et al., 2014). With every 10 m corresponding to 1 atmosphere of pressure, this places the maximum known pressures in the deep between 1092.4 and 1103.4 atm (110.7–111.8 MPa). Notable is that several other deep-sea trenches—Tonga, Philippines, Kuril-Kamchatka, and Kermadec—in the Pacific Ocean also reach depths >10,000 m. In the Atlantic Ocean, the two deepest trenches are the Puerto Rico Trench at 8800 m and the South Sandwich Trench at 8428 m.

The prevalence of prokaryotic and eukaryotic life through the range of high pressures in the deep oceans suggests that organisms can easily adapt to these conditions. Piezophiles and hyperpiezophiles that preferentially reproduce at high pressures (10–50 and 50 MPa, respectively) represent a broad set of microbes that occur across the expansive abyssal (∼5000 m) and hadal (>5000 m) depths. Even at the extreme pressures (>110 MPa) in the Challenger Deep, microbial life flourishes and differs from the background abyss (Tarn et al., 2016; Xu et al., 2018). At the microbial level, pressure generally increases piezophile activities in oxic deep sea and in anoxic vents and subseafloor conditions (Picard and Daniel, 2013).

Earlier explorations of the Challenger Deep microbial community did not recover expected piezophilic bacteria (Takami et al., 1997). However, subsequent work found piezophilic species of Colwellia strain KT27, Colwellia piezophila, and Moritella abyssi (Tarn et al., 2016) with higher abundances than those found in the Puerto Rico Trench. In addition, a marked decrease in chemolithoautotrophs occurs at trench depths, replaced by heterotrophic bacteria potentially reflecting the concentration of organic material that occurs within trenches (Nunoura et al., 2018). Indeed, high rates of microbial carbon turnover occur in the Challenger Deep in response to this elevated deposition of organic material (Glud et al., 2013). Much of the microbial community in the Challenger Deep is dominated by Gammaproteobacteria and particularly Oceanospirillales (Tarn et al., 2016; Nunoura et al., 2018), with strong representations of Marinimicrobia, Pelagibacter, and the cyanobacterium Prochlorococcus as well (Tarn et al., 2016). High counts of Betaproteobacteria, including Alcanivorax and Thiobacteraceae, were also found in the Challenger Deep. At >10,400 m in the Challenger Deep, hydrocarbon-degrading bacteria within Oceanospirillales including Oleibacter, Thalassolituus, and Alcanivorax also reach densities higher than anywhere on the Earth (Liu et al., 2019). Both what the hydrocarbon source is and why the prominent shift is at 10,400 m remain unanswered (Liu et al., 2019).

Several large multicellular organisms have been found at extreme depths and pressures and can be characterized as primarily hadal and trench species (Wolff, 1970; Jamieson et al., 2010). Endemism of species occurring greater than 7000 m is 74% of surveyed species (Wolff, 1970), implying a strong evolutionary pressure and the unique adaptations required to exist at extreme depths. One perhaps critical limitation is that pressure limits larval phases and may prevent many species from inhabiting high-pressure environments (Young et al., 1997). Wolff (1970) provides an in-depth review of the maximum depths reached by benthic organisms across several groups, and readers are referred to this publication. Among invertebrates, several reach trench depths, including the sea cucumber Myriotrochus bruuni at 10,710 in the Mariana Trench, the extremely abundant amphipod Hirondellea gigas at 10,897 m in the Challenger Deep, the isopod Macrostylis species at 10,710 in the Mariana Trench, the polychaete genus Macellicephaloides at 10,700 m depth, and the sea anemone family Galatheanthemidae at 10,700 m depth (Wolff, 1970; Kobayashi et al., 2012). The deepest known vertebrate is the hadal snailfish Pseudoliparis swirei found at 7966 m (Gerringer et al., 2017). It is also worth noting several species occur at great numbers at extreme depths, for example, 1500 individuals of a Siboglinidae at 10,000 m in the Kuril-Kamchatka Trench (Wolff, 1970).

Extreme pressure requires numerous biochemical and genomic adaptations (Somero, 1992; Simonato et al., 2006). However, the current understanding is that the viability of piezosensitive microorganisms is severely decreased at pressures around ∼100–150 MPa (Picard and Daniel, 2013). Extreme pressure results in a tighter packing of the phospholipids in the cell membrane, which in turn lowers cellular permeability (Simonato et al., 2006). Many deep-sea organisms increase the percentage of unsaturated fatty acids (Somero, 1992; Simonato et al., 2006). The double covalent bond between adjacent carbons in an unsaturated fatty acid leads to a “kink” in the tails of the molecule, allowing for looser packing of the cellular membrane (Somero, 1992). Two piezophilic bacteria isolated from the Mariana Trench, Shewanella sp. strain DB21MT-2 and Moritella sp. strain DB21MT-5, contain high proportions of the monounsaturated fatty acids (Simonato et al., 2006). A breakdown of biological function due to the dissociation of protein complexes occurs at 200–300 MPa (Hazael et al., 2016). Increased pressures also select for different enzymes, with selection for rigidity to counteract pressure and the resulting warping of proteins. Deep-sea proteins contain increased hydrogen and disulfide bonds between different subunits, which dictate structure, to minimize changes in shape to do pressure (Somero, 1992). The presence of organic solutes, “piezolytes,” can also counteract the pressures on proteins (Yancey et al., 2014). Trimethylamine N-oxide (TMAO) has received the most attention and is known to stabilize proteins against temperature and the toxic effects of urea (Yancey et al., 2014). TMAO has been shown to stabilize proteins against hydrostatic pressure in a broad set of organisms from yeast to mammals (Yancey et al., 2014), and the concentration of TMAO limits species maximum depth of occurrence in the deep sea (Yancey et al., 2014).

Pressure, along with temperature, also impacts the “carbonate compensation depth” (CCD), the depth at which calcium carbonate goes in and out of solution equally and beyond which dissolution is greater. The CCD thus acts as a filter on species with calcium carbonate skeletons, for example, species of foraminiferans, corals, crustaceans, molluscs, and echinoderms (Carney, 2005). In echinoderms, ossified taxa such as ophiuroids and echinoids are replaced by soft-bodied taxa such as holothurians, with increasing depth (Carney, 2005). For gastropods, the CCD limits their depth distributions favoring species with increased overlapping of shell whorls, less ornamentation, and decreased surface area-to-volume ratios that limit exposure of the shell to dissolution (McClain et al., 2004). Notable is that while gastropods have been found at 7703 m in the Japan Trench, they were soft-shelled species (Wolff, 1970).

3.2. Extreme toxicity

Toxic materials in the deep sea can be broadly classified as hydrocarbons, metals (e.g., iron, manganese), or nonmetals (e.g., sulfide). Anthropogenic stressors such as spills from oil extraction and deep-sea mining increase local toxicity in areas where life is not adapted to it (Knap et al., 2017). However, many natural areas exist where evolutionary innovation has allowed organisms to adapt to living, and even thriving, in toxic conditions. Natural toxicity is concentrated in three environments: hydrothermal vents, cold seeps, and brine pools. At hydrothermal vents, fluids emitted contain high loads of sulfur and toxic metals, leached from the mineral-rich chimney structures (Van Dover, 2000). These toxic materials have a sphere of influence larger than the vent itself, with metal deposits reaching the sediment around each vent, drawing recent interest for mining activities (Aldhous, 2011). The most comprehensively studied hydrothermal vents occur along the Mid-Atlantic Rise, and the Rainbow vent site is the most toxic of these, where low pH and high chlorine levels and temperature create an environment rich in metals (Douville et al., 2002). Here, iron levels reach 24,000 μM, over four times higher than the concentrations of iron at any other known vent, and manganese levels are an order of magnitude higher than any other vent (2250 μM; Douville et al., 2002). Cold seeps emit methane from beneath the sediments into the overlying water at concentrations orders of magnitude higher than the surrounding sediments, up to ∼80 mM (Wankel et al., 2010; Lapham et al., 2013). These environments are also associated with high levels of sulfide due to high rates of sulfate reduction by heterotrophic and autotrophic sulfate-reducing microorganisms (Bowles et al., 2011). Brine pools vary in metal and sulfide concentrations, primarily determined by their type, with those that have hydrothermal influence containing the highest metal concentrations (Antunes et al., 2020). In the Gulf of Mexico brine pools, methane constitutes over 99% of the total hydrocarbons present (Joye et al., 2005). The Mediterranean brine pool Urania is the most sulfidic water body on earth, with sulfide concentrations reaching 16 mM (Antunes et al., 2020). However, it is important to note that gassy and oily sediments, typically found at cold seeps, can often lead to pore water sulfide levels >20 mM (Joye et al., 2005; Bowles et al., 2016).

Archaea and bacteria thrive in these toxic habitats, both alone and in symbiosis with animal hosts. Large swaths of free-living bacterial mats, typically Beggiatoa, Thioploca, and Thiomargarita are common to brine pools and other cold seep sites (Joye et al., 2005). These microbes, which are large enough to be seen with the naked eye, thrive at the interface of oxygenated and sulfidic sediments by oxidizing sulfide with nitrate. At hydrothermal vents, the genus Sulfurimonas is abundant, as these organisms oxidize sulfur associated with sulfide release from vent fluids (Mino et al., 2017; McNichol et al., 2018). In general, bacteria isolated from hydrothermal vent fluids exhibit tolerance to high concentrations of mercury cations and the ability to reduce this toxic material to elemental mercury (Vetriani et al., 2005). Hydrothermal vent archaea make use of sulfide to generate metal–sulfide complexes, which decrease the bioavailability of the metals and in turn create a natural buffer against high metal concentrations (Edgcomb et al., 2004). In addition to free-living prokaryotes, methanotrophic and chemosynthesizing symbiotic bacteria form characteristic symbiotic relationships with a wide variety of invertebrates common to these habitats, including tube worms, mussels, sponges, flatworms, clams, snails, shrimp, and crabs (Dubilier et al., 2008).

Some of the most iconic animals to live in toxic conditions are the mussels of the Bathymodiolus genus—pervasive around hydrothermal vents, cold seeps, and brine pools—that have developed physiological adaptations to deal with, and even thrive in, a broad range of toxic habitats. For example, Bathymodiolus individuals at one Gulf of Mexico brine pool, which form a band around the pool from the edge extending outward 3–6 m, experience the best growth and fitness parameters within 1 m of the edge of the brine pool despite high methane concentrations of >200 μM (Smith et al., 2008). Bathymodiolus mussels collected from hydrothermal vent fields and cold seeps in the western Pacific possess increased metal concentrations in their tissues (Zhou et al., 2020), which can lead to physiological problems such as inhibition of DNA repair mechanisms (Hartwig, 1998), and oxidative stress (Zhou et al., 2020). To combat these issues, the mussels reproduce early, taking advantage of a transient capacity to repair damaged DNA while they are small (Pruski and Dixon, 2003), possess high levels of both enzymatic antioxidants, particularly in the gill tissues where contact with metal-rich fluids is the most frequent, and use several nonenzymatic agents that further assist in metal detoxification (Zhou et al., 2020). Some evidence suggests that the antioxidant defense enzymes in the gill tissues may be supplemented by gill symbionts (Company et al., 2006).

Other organisms have also developed strategies to combat toxicity. The giant tube worms Riftia pachyptila are dominant in diffuse flow regimes of hydrothermal vent fields (Childress et al., 1991), which vary substantially in chemistry but can reach hydrogen sulfide concentrations of >250 μmol kg⁻¹ (Le Bris et al., 2006). These tube worms are entirely reliant on chemosynthesis by symbionts for food (Cavanaugh et al., 1981), a process that requires a considerable amount of sulfide, yet this sulfide is detrimental to the worm's aerobic pathways (Girguis and Childress, 1998). To strike a balance between these two important needs, R. pachyptila mitigates the toxicity of H₂S by quickly converting it to HS⁻ within the gill epithelia and binding it to hemoglobins (Girguis and Childress, 2006). The symbionts of R. pachyptila also appear to confer a superoxide dismutase (SOD) to combat oxidative stress from metal exposure at hydrothermal vent fields, as the specific type of SOD observed in the trophosome tissue of the tubeworm was previously unknown from animal tissue (Blum and Fridovich, 1984). Other examples of adaptation to toxicity include the Alvinocaris longirostris shrimp, which live among mussel beds at the base of active hydrothermal vents (Watabe and Miyake, 2000). The transcriptome of these shrimp possesses genes related to metal detoxification, sulfur metabolism, and defense against oxidative stress, suggesting that selection and duplication or high expression of particular genes may be integral to the adaptation process (Sun et al., 2018). Even species that are transient to these habitat types appear to have adaptations for dealing with metal toxicity; for example, Rimicaris sp. shrimp exhibit induction of metallothionein, a metal-binding protein used in detoxification or storage, in response to metal exposure (Auguste et al., 2016).

3.3. High salinity and brines

The highest salinities in the deep sea are found in brine pools; most abundant in the Gulf of Mexico, the Mediterranean Sea, and the Red Sea. In brine pools, excess salt is not the only physiological hurdle for life to overcome—these environments are often characterized by a combination of environmental extremes, including high salinity, low oxygen, low or high temperature, and relatively high concentrations of toxic metals. These phenomena are relatively newly discovered and much remains unknown about them, including the maximum salinity. Some brine pools are vertically stratified; salinity and temperature increase with depth, while oxygen decreases (Bougouffa et al., 2013; Pachiadaki et al., 2014). Compared with the average seawater salinity of ∼35 psu, the salinity of two Red Sea brine pool sites, Atlantis II and Discovery, increases from 50 psu at the interface of the brine and overlying seawater to nearly 250 psu at depth (Bougouffa et al., 2013). A third Red Sea brine pool, Kebrit, has salinity ranging 40.6–242 psu (Vestheim and Kaartvedt, 2016), and a fourth unnamed Red Sea pool ranged from 43 to 96 psu (Wang et al., 2014). In the Mediterranean brine pool Thetis, salinity increases from 35 psu at the brine pool surface to over 340 psu at depth (Pachiadaki et al., 2014). Similar patterns and salinities have been observed in the Gulf of Mexico, where salinities can reach as high as ∼140 psu (e.g., Atwater Valley and Garden Banks) (Bowles et al., 2016).

The environmental conditions of brine pools are toxic to large metazoans. However, microbial life persists with intense activity (DasSarma, 2006; Joye et al., 2009). Many new species and at least one new bacterial phylum (Garrity et al., 2001) have been discovered from these extreme environments. In the Red Sea, brine pools Atlantis II and Discovery host cell densities within the brine twelve times higher than that of the overlying seawater, reaching maxima of 54.5 ± 15.45 × 10⁴ cells/mL and 28.3 ± 10.41 × 10⁴ cells/mL in Discovery and Atlantis II, respectively (Bougouffa et al., 2013). In addition, the brine supported a higher diversity than the overlying waters, with the deepest layers of the brine supporting the highest diversity of both bacteria and archaea species (Bougouffa et al., 2013). The stratified layers of brine host unique communities of bacteria, with distinctive taxa dominating in each layer (Bougouffa et al., 2013; Pachiadaki et al., 2014). In the Mediterranean Thetis brine pool, bacteria and archaea dominated, but fungi, dinoflagellates, and ciliates were also present, indicating that even eukaryotic life can exist despite the challenges (Pachiadaki et al., 2014). Fungi were also present, particularly in the sediments at the bottom of the pool, in an unnamed, relatively shallow brine pool in the Red Sea (Vestheim and Kaartvedt, 2016).

The conglomeration of extreme conditions represented by brine pools certainly requires evolutionary ingenuity. Indeed, microbial communities that live within brine are significantly different than those that live in overlying deep waters (Bougouffa et al., 2013; Pachiadaki et al., 2014; Wang et al., 2014). Bougouffa et al. (2013) suggest that bacteria and archaea are able to persist in these environments by activating stress responses, detoxification, and repair mechanisms. Evidence from the Thetis brine pool suggests that microorganisms deal with osmotic stress through osmolyte synthesis, especially glutamine, or by adapting ion transport strategies (Pachiadaki et al., 2014). Some halophiles also bring in potassium and glycine-betaine to offset high environmental salt levels (Rothschild and Mancinelli, 2001). Many groups of halophilic prokaryotes have apparent near-global distribution, which raises the question of dispersal adaptations or potential long-term periods of dormancy and preservation within salt crystals (Antunes et al., 2020). In addition, salinities of the deepest layers of brine pools are too high to allow for autotrophy because of low water potential, and thus, prokaryotes must acquire energy by a yet unknown pathway (Pachiadaki et al., 2014).

3.4. Extremes of acidity or alkalinity

The average pH of seawater is ∼8.1, although at various sites along the seafloor, including some previously mentioned, higher and lower pH conditions can exist. Microorganisms have adapted to thrive in acidic and alkaline conditions. In fact, microbial activity has been observed at pHs ranging from 0 to 14 (Schleper et al., 1996; Yang et al., 2008). Specific adaptations of prokaryotes for acidic and alkaline conditions exist with commonalities being that they both try to maintain a more neutral internal pH and have cellular membrane lipid (Horikoshi, 2006; Baker-Austin and Dopson, 2007; Siliakus et al., 2017). There are very specific mechanisms that acidophilic and alkalophilic microorganisms, respectively, utilize to protect and maintain viability, which are beyond the scope of this review (Lund et al., 2002; Krulwich et al., 2007). Within the deep-sea environment, the ranges of pH from highly acidic to highly alkaline are found at traditional hydrothermal vents derived from magma heating and off-axis sea mount vents that exist as a result of the reaction of seawater with mantle rock. Hydrothermal vents associated with magma, those often found at lithospheric plate boundaries, inject hot fluids into the seawater, which results in very acidic water at the interface of hydrothermal fluids and seawater (Ding et al., 2005). The pH at these sites varies but can be as low as ∼3 (Seyfried et al., 2003; Ding et al., 2005). These pH gradients occurring because of leaky hydrothermal chimneys may also be important in complex carbon molecule production and the origins of life (Ooka et al., 2019). This is a sharp contrast to the alkaline vents of seamounts, most notably the Lost City Hydrothermal Field, where relatively cooler hydrothermal fluid is produced by the interaction of olivine with seawater, that is, the serpentinization reaction. This reaction also results in the production of the basic hydrothermal fluids with pHs 9–10 (Kelley et al., 2001). Indeed, at both vent types with acidic and alkali extremes, viable organisms have been observed (Reysenbach et al., 2006; Lang and Brazelton, 2020). The Lost City site is particularly interesting regarding astrobiology as Glein et al. (2015) predict that the ocean of Enceladus is alkaline.