Oxygen isotope ratios of PO₄: An inorganic indicator of enzymatic activity and P metabolism and a new biomarker in the search for life

Abstract

The distinctive relations between biological activity and isotopic effect recorded in biomarkers (e.g., carbon and sulfur isotope ratios) have allowed scientists to suggest that life originated on this planet nearly 3.8 billion years ago. The existence of life on other planets may be similarly identified by geochemical biomarkers, including the oxygen isotope ratio of phosphate (δ¹⁸O_p) presented here. At low near-surface temperatures, the exchange of oxygen isotopes between phosphate and water requires enzymatic catalysis. Because enzymes are indicative of cellular activity, the demonstration of enzyme-catalyzed PO₄–H₂O exchange is indicative of the presence of life. Results of laboratory experiments are presented that clearly show that δ¹⁸O_P values of inorganic phosphate can be used to detect enzymatic activity and microbial metabolism of phosphate. Applications of δ¹⁸O_p as a biomarker are presented for two Earth environments relevant to the search for extraterrestrial life: a shallow groundwater reservoir and a marine hydrothermal vent system. With the development of in situ analytical techniques and future planned sample return strategies, δ¹⁸O_p may provide an important biosignature of the presence of life in extraterrestrial systems such as that on Mars.

The basic ingredients required for life include a source of chemical building blocks, energy, and liquid water (1). Mars has possessed all of these attributes in its history and may even have extensive reservoirs of subsurface liquid water today (2, 3). Phosphorus is essential to life on Earth and, therefore, an obvious candidate for a chemical tracer of the presence of life in extraterrestrial systems. Results from recent Mars Pathfinder surveys show relatively high concentrations of P (≈0.4 wt % P₂O₅) in Mars surface sediments that are thought to be derived from dissolution of apatite by aqueous fluids percolating through the surface layers of Mars and later enriched by evaporation.¶ It has been suggested that anomalies in phosphorus concentrations and P/Th ratios in sediments might be useful tracers of extinct life on Mars (4). The occurrence of phosphate, particularly in association with organic carbon, has been interpreted as the remains of once living organisms in ancient sedimentary strata (5, 6). Here, we demonstrate that the unique chemical and isotopic properties of PO₄ make oxygen isotope ratios of phosphate (δ¹⁸O_P) an ideal signature of the presence of enzymatic activity and, thus, life.

Stable isotope ratios can provide diagnostic signatures of biological activity because of the large and characteristic isotopic fractionations associated with many metabolic reactions (e.g., bacterial sulfate reduction and photosynthesis). As a result, the systematics of stable isotope fractionation for key bioelements (C, N, and S) have been studied extensively, and the isotopic ratios of these elements have been widely used as biosignatures in both terrestrial and extraterrestrial samples. An analogous isotopic biomarker system is not available for P, because, unlike C, N, and S, it has only one stable isotope. Phosphorus is also distinct from other major bioelements in that it occurs primarily in one oxidation state (+5), and in one major form, orthophosphate (PO₄). δ¹⁸O_P values have been used primarily as a paleotemperature proxy recorded in biogenic apatite minerals found in teeth, bones, fish scales, and shells (7). The PO₄ radical is fairly inert to chemical redox reactions and to oxygen isotope exchange at low temperature (<80°C).

Enzymes are required for catalysis of oxygen isotope exchange between PO₄ and water at low temperature. They are often invoked as the cause of diagenetic alteration of phosphate minerals leading to the loss of integrity of original δ¹⁸O_P values. Detailed studies of enzyme-catalyzed PO₄-H₂O exchange reactions (8–10) demonstrate that one can exploit this feature to assess δ¹⁸O_P values in a fundamentally different way: as an indicator of the presence of enzymatic activity and, hence, living organisms. Few investigators have examined oxygen isotope compositions of dissolved phosphate and nonbiogenic phosphate in modern sediments/soils, and, to our knowledge, no one has attempted to make use of the unique dependence of PO₄–H₂O oxygen isotope exchange on enzyme activity specifically as a biomarker. Here, we demonstrate the application of δ¹⁸O_p as a biomarker in two terrestrial systems that serve as analogues for extraterrestrial planetary environments highly targeted in the search for life: a groundwater aquifer, and a marine hydrothermal vent system. Continued advances toward in situ sampling and analytical techniques, as well as sample-return missions planned for Mars, make δ¹⁸O analysis of phosphate and water in extraterrestrial materials plausible in the near future.

Method Development

Analytical Methods.

Oxygen isotope ratios of PO₄ reported in this study were determined by conversion of inorganic PO₄ extracted from groundwater or sediments into silver phosphate. The silver phosphate was then converted to CO₂ after the method of O'Neil et al. (11) or by conventional fluorination by using ClF₃. Water samples were analyzed by the standard method of CO₂–H₂O equilibration (12). Isotopic analyses were carried out in the stable isotope facilities at the University of Michigan (ISOLAB) and Yale University. Precision of the isotopic analyses was ±0.1‰ (1σ) for water and ±0.2‰ for phosphate.

All oxygen isotope data are reported in the standard delta notation in per mil (‰) relative to the SMOW (standard mean ocean water) international reference standard. The fractionation factor, α, between PO₄ (p) and water (w) is defined as α_p-w = (1 + δ¹⁸O_P/1,000)/(1 + δ¹⁸O_W/1,000).

Microbial growth and phosphate metabolism studies were carried out as described in the legends of associated figures and data tables.

Calibration of the δ¹⁸O_P Biomarker.

To employ this potential biomarker, we must first consider the processes recorded by inorganic PO₄ dissolved in natural waters and associated with sediments. Reactions of PO₄ in natural waters, sediments, and soils are largely under biological control because of the importance of P as a nutrient for microorganisms and plants. There are very few studies of the oxygen isotope systematics of dissolved PO₄ and P cycling in natural waters and sediments (13, ‖). However, oxygen isotope compositions of PO₄ precipitated by living organisms in the form of biogenic apatite minerals (teeth, shells, and bone) have been studied extensively (e.g., refs. 14–16). Biogenic phosphate is equilibrated with body fluids of an organism as a result of intracellular enzymatic reactions (e.g., ATP utilization and phosphorylation/dephosphorylation). The temperature dependence of oxygen isotope fractionation between apatite and water has been determined both empirically and in laboratory experiments (17–20). The most widely cited relation is the empirical equation of Longinelli and Nuti, t (°C) = 111.4 − 4.3[δ¹⁸O_P − δ¹⁸O_W], which is generally assumed to reflect equilibrium isotopic fractionation (17).

In contrast to biogenic phosphates, where the exchange of PO₄ with body fluids and subsequent precipitation as apatite occur in a closed system, the metabolism of P in natural waters is mediated largely by microorganisms in a relatively open system. This phenomenon occurs because microbial metabolism of P compounds (inorganic and organic) involves the action of extracellular phosphate-scavenging enzymes and intracellular–extracellular PO₄ exchange processes (ion transport proteins), resulting in the release of inorganic PO₄ into the extracellular environment (21). Many of these enzymatic reactions involving PO₄ are dominated by hydrolytic cleavage reactions that facilitate the incorporation of water (from the medium) into PO₄. These reactions promote significant PO₄–H₂O oxygen isotope exchange that is reflected in the δ¹⁸O value of dissolved PO₄. Biochemists have exploited this property for decades in elegant studies of phosphoenzyme reaction kinetics and mechanisms by using heavily labeled (e.g., >90% ¹⁸O) water and P compounds (22, 23). The capacity of δ¹⁸O_P as a biomarker at natural ¹⁸O abundance levels also results from its ability to record enzymatic activity, and from the lack of significant PO₄–H₂O oxygen isotope exchange by other abiotic geochemical processes involving P. For example, the precipitation of dissolved phosphate as apatite, important during sediment diagenesis, does not involve exchange of PO₄ oxygen isotopes and results in only small dissolved PO₄-apatite fractionations (≈1‰; R.E.B. and Y. Liang, unpublished data).

The low dissolved PO₄ concentrations found in most natural systems, coupled with intense biological recycling and turnover characteristic of P (24), should ensure complete oxygen isotope exchange between dissolved PO₄ and ambient water. Thus, oxygen isotope ratios of dissolved PO₄ in natural waters and associated with sediments should record the effects of microbial P metabolism and microbial enzyme activity at the ambient environmental water and temperature conditions.

In summary, δ¹⁸O_P values can be used to detect enzymatic activity in two ways: (i) by recording exchange between PO₄ and ambient water, and (ii) by recording ambient temperature. Both result from the process of PO₄–H₂O oxygen isotope exchange. At low near-surface temperatures, PO₄–H₂O exchange is catalyzed by the action of enzymes. Results from detailed laboratory studies of oxygen isotope fractionation associated with microbial metabolism of P compounds and enzyme catalyzed PO₄–H₂O exchange reactions provide a context for the interpretation of δ¹⁸O_P values and for the detection of enzymatic activity in natural systems.

Relation between δ¹⁸O_P, δ¹⁸O_W, and Temperature.

Extensive oxygen isotope exchange between water and dissolved PO₄ is demonstrated in laboratory experiments where bioavailable P compounds (inorganic orthophosphate and organic phosphoesters) were metabolized by microorganisms (Fig. 1; Table 1) or degraded by purified cell-free phosphoenzymes (Fig. 2; Table 2; refs. 8 and 9). This exchange occurs even at dissolved PO₄ concentrations that are more than 1,000 times those typical of natural systems. Phosphate–water exchange is reflected in a strong positive correlation between the δ¹⁸O value of water and the δ¹⁸O value of dissolved PO₄ (Figs. 1 and 2). Experiments with bacteria also show that microbial metabolism of P is characterized by equilibrium oxygen isotope fractionations that vary with temperature in a manner similar to that observed for biogenic apatites (Fig. 3; Table 3). Similar results were obtained in experiments on the temperature dependence of PO₄–H₂O oxygen–isotope exchange catalyzed by cell-free, purified enzymes (R.E.B., J. R. O'Neil, and G. A. Garcia, unpublished data), demonstrating that catalytic molecules that are derived from, or possibly precursors of, complex cells may also leave behind a distinctive geochemical signature of biological activity.

Figure 1.

Microbial P metabolism experiments. Mixed (from soils, rivers) and pure strains of microorganisms [e.g., Klebsiella, Bacillus, Aquaspirillum) were grown on representative bioavailable P compounds (inorganic orthophosphate (P_i), organic P monoesters (glucose-PO₄) and diesters (RNA)] to study PO₄–H₂O oxygen isotope exchange during microbial metabolism of P. Microbial metabolism of P leads to a positive correlation between δ¹⁸O of water and δ¹⁸O of dissolved PO₄ in the medium, for both inorganic and organic P substrates. Plots show representative data from 25°C and include data from Blake et al. (8, 9).

Table 1.

Oxygen isotope compositions of water and dissolved inorganic PO₄ derived from various P sources and metabolized by microbial cell cultures at 25°C and plotted in Fig. 1

P source	δ18Ow	δ18Op
5 mM Pi	99	82.6
10 mM Pi	−17.7	7.3
	−6.0	18.0
	12.6	35.6
RNA	−6.1	13.0
	−6.5	12.8
	−19.3	3.6
	−8.0	11.1
	10.4	21.7
Glucose-1-PO4	−18.2	4.1
	−6.7	12.1
	42.2	57.9

Figure 2.

Purified cell-free enzyme catalyzed PO₄–H₂O oxygen isotope exchange. Positive correlation between δ¹⁸O of dissolved inorganic PO₄ and water resulting from PO₄–H₂O exchange catalyzed by hydrolytic phosphoenzymes inorganic pyrophosphatase (PPiase) and alkaline phosphatase (APase). (a) PPiase exchange experiments at 22°C and 30°C in water with δ¹⁸O of −19.7‰ to 96.4‰. The initial δ¹⁸O_p value of dissolved inorganic PO₄ before exchange was 13.5‰. (b) APase exchange experiments in −19.4‰ and 44.1‰ water at 37°C using glycero-2-phosphate as an organic-P substrate (P source).

Table 2.

Oxygen isotope compositions of water and dissolved inorganic PO₄ derived from reaction of organic-P compounds and inorganic PO₄ with purified, cell-free phosphatase (hydrolytic) enzymes

P source	δ18Ow	δ18Op
Alkaline phosphatase at 37°C
Glycero-2-PO4	−19.4	−5.2
	44.1	13.1
Inorganic pyrophosphatase at 22–30°C
20 mM Pi	−19.7	3.6
10 mM Pi	14.2	33.7
10 mM Pi	−19.5	1.8
10 mM Pi	96.4	105.8

Data plotted in Fig. 2. 

Figure 3.

Temperature dependence of PO₄–H₂O oxygen isotope exchange accompanying microbial metabolism of P compounds. The PO₄–H₂O fractionations produced by bacteria agree well with fractionations determined from biogenic apatites. Biogenic apatite data are from Longinelli and Nuti (18) and are the basis for the PO₄–H₂O thermometry equation: t (°C) = 111.4 − 4.3[δ¹⁸O_P − δ¹⁸O_W] (18). Microbial phosphate data from Blake et al. (8, 9).

Table 3.

Oxygen isotope ratios of water and dissolved inorganic phosphate liberated during microbial degradation of organophosphorus compounds (RNA) at 20 to 35°C

T, °C	δ18Ow	δ18Op	103 lnα
20	−6.3	13.4	19.63
25	−6.5	12.7	19.14
25	−6.1	13	19.03
30	−6.2	12.2	18.35
35	−6.2	11.5	17.65
35	−6.6	10.8	17.36
35	−5.5	11.7	17.15
25	−8	11.6	19.57
35	−0.5	16.7	17.06

Data plotted in Fig. 3. 

The above results indicate that microbial metabolism of P should result in oxygen isotope exchange between dissolved PO₄ and ambient water, regardless of the nature (organic or inorganic) or initial δ¹⁸O value of the P source. Further, such metabolized PO₄ should record the temperature of equilibrium exchange according to the same relation used for oxygen isotope thermometry of biogenic apatites. Therefore, in conjunction with measured or inferred temperature and δ¹⁸O_w values, the δ¹⁸O value of dissolved PO₄ and sediment-associated PO₄ can be used to detect enzymatic activity and, thus, infer the presence of life.

Application of the δ¹⁸O_P Biomarker

Dissolved PO₄ in Groundwater—Correlation Between δ¹⁸O_w and δ¹⁸O_P.

Oxygen isotope ratios of dissolved PO₄ from a shallow glacial outwash aquifer in Cape Cod, MA were measured in a study of microbial turnover of PO₄ in groundwater.^¶ Dissolved phosphate concentrations (30–108 μM) reach levels that are 1 to 2 orders of magnitude higher than those in typical aquatic and marine systems because of contamination of the ground water by sewage. Phosphate within the contaminant plume is not fully equilibrated with ground water at measured ambient aquifer temperatures (Table 4). Complete metabolic turnover and exchange of the entire phosphate pool with water is probably limited by low dissolved organic carbon concentrations in the aquifer. At dissolved PO₄ concentrations more typical of uncontaminated systems (<5 μM), it is expected that the entire phosphate pool will be metabolized and more fully equilibrated with water.

Table 4.

δ¹⁸O of dissolved inorganic PO₄ and groundwater in the Cape Cod aquifer

δ18Ow	δ18Op	T, °C
−8.0	18.4	9.5
−7.8	19.3	7.7
−7.7	19.1	7.7
−8.4	16.4	8.2
−7.9	18.4	13.7

δ¹⁸O values of dissolved PO₄ in the groundwater show clear evidence of phosphate metabolism. There is a positive correlation between the δ¹⁸O of water and the δ¹⁸O of dissolved PO₄ (Fig. 4). This correlation suggests that extensive oxygen isotope exchange has occurred between the water and dissolved PO₄. Importantly, temperatures measured within the aquifer, approximately 8°C to 14°C, are much too low for the observed PO₄–H₂O exchange to have occurred by abiotic, nonenzymatic reactions. Therefore, enzymatic activity and metabolism of dissolved inorganic PO₄ in the groundwater is indicated.

Figure 4.

Comparison between the δ¹⁸O of dissolved PO₄ and groundwater from the Cape Cod aquifer. Positive correlation between δ¹⁸O_P and δ¹⁸O_W at low temperature (8°C to 14°C) indicates the presence of enzyme catalyzed PO₄–H₂O oxygen isotope exchange and active microbial metabolism of P within the aquifer.

δ¹⁸O of Dissolved Inorganic PO₄ Associated with Sediments: Fe Oxide Deposits in a Marine Hydrothermal System.

Currently, it is widely held that life on Earth may have originated in hydrothermal systems (25). Protected from intense UV irradiation by overlying water, the first biomolecules and primitive cellular life may have evolved in deep-sea vent environments by using the energy supplied by hydrothermal activity and reduced forms of sulfur and hydrogen (25). Similar hydrothermal systems may have been active over 3 billion years ago on Mars (26). Phosphate in active and fossil hydrothermal systems may provide evidence of metabolic activity.

Dissolved inorganic PO₄ can become strongly associated with sediments, especially iron oxides, via sorption or precipitation processes (27). Hydrothermal iron oxide deposits are produced by oxidation of erupting hydrothermal fluids rich in reduced iron, by ambient oxygenated seawater (27, 28) and in some cases, may involve the action of iron-oxidizing bacteria (29, 30). Iron oxides are very efficient scavengers and natural concentrators of dissolved PO₄ (27, 31).

The δ¹⁸O_P of inorganic PO₄ associated with Fe oxide deposits (Fe–PO₄) was analyzed from two diffuse flow hydrothermal systems at Larson's Seamounts located near 21°N along the East Pacific Rise (EPR) and from Seamount No. 5, located 86 km east of the EPR near 13°N (refs. 29 and 32, and Table 5). These hydrothermal systems are located off-axis of the main EPR spreading center. One sampling site at Red Seamount was experiencing active venting at the time of sampling whereas the other sites were inactive and comprised fossil hydrothermal deposits.

Table 5.

δ¹⁸O values of PO₄ associated with hydrothermal Fe-oxide deposits

Sampling location	δ18Op	T, °C calc.*
Green Seamount	25.6 (n = 4)	1.3
Seamount no. 5	25.1 (n = 3)	3.5
Red Seamount	11.9 (n = 4)	60.2
Red Seamount	23.4 (n = 7)	10.8

^*Temperatures calculated using the Longinelli and Nuti equation (17). 

Evidence for enzymatic activity and microbial metabolism of the PO₄ associated with these Fe oxide deposits comes from the temperatures recorded in PO₄ oxygen isotope ratios. Hydrothermal systems provide a variable temperature structure that can be used to detect PO₄–H₂O oxygen isotope exchange in an otherwise constant δ¹⁸O_w environment of ambient seawater. Iron oxides at Red Seamount are believed to have formed by precipitation from actively venting hydrothermal fluids. At Green Seamount, the ocherous, Fe oxide sediments are interpreted to be secondary alteration products formed by oxidation of iron sulfide minerals precipitated from hydrothermal fluids (29, 32, 33).

δ¹⁸O values of PO₄ associated with iron oxide deposits collected from Red Seamount record temperatures that are very warm relative to ambient seawater (2°C–5°C; Table 5). One sample taken from the upper 40 cm of an actively venting orifice yielded a temperature of 10.8°C that closely matches measured vent temperatures of 10°C–15°C (29). δ¹⁸O_p temperatures were calculated by using the PO₄–H₂O temperature equation (17) and assume a δ¹⁸O of seawater of 0‰. The recording of temperatures as low as 10.8°C in δ¹⁸O_P values strongly suggests that the PO₄ has been metabolized and has undergone oxygen isotope exchange with seawater by enzyme-catalyzed reactions at ambient temperatures.

Temperatures calculated from δ¹⁸O values of Fe-PO₄ at another site at Red Seamount that was not actively venting at the time of sampling averaged 60°C. This higher temperature suggests the venting of warm temperature fluids at this site in the past. Further evidence for higher temperature processes in the past at Red Seamount is provided by the presence of iron-rich talc deposits and from temperatures calculated from δ¹⁸O values of nontronite deposits (32). A temperature of 60°C, however, is still too low to promote significant PO₄–H₂O oxygen isotope exchange by inorganic/abiotic means. Thus, the warm temperatures recorded in δ¹⁸O_p values of these Fe-PO₄ deposits also suggest the presence of enzymatic activity and microbial metabolism of the PO₄.

Oxygen isotope exchange between PO₄ and water may also have occurred deeper within the sediments or in the underlying basaltic basement rocks at higher temperatures, by the action of subsurface bacteria. Subsurface vent microbial habitats have been reported from Loihi Seamount, where thermophilic bacteria with high metabolic activity were observed at temperatures of 60°C (30).

The mechanism proposed for exchange of Fe-PO₄ with water is microbial uptake and metabolism of the phosphate. This mechanism requires that PO₄ bound to Fe oxides be biologically available. To assess the bioavailability of Fe oxide-associated PO₄, separate aliquots of the Fe oxide sediments analyzed for δ¹⁸O_p were used in microbial growth experiments as a sole source of P. The two strains of bacteria tested in these experiments, Acinetobacter ADP1 and Shewenella MR-1, were able to grow readily on minimal media with Fe-PO₄ sediments supplied as the sole source of P (Fig. 5). No dissolved PO₄ was detected in the Fe-PO₄ growth media initially or over the course of the growth experiments, as determined by periodic extraction and analysis of growth media for dissolved PO_4. Under the oxic conditions of the growth experiments and also of natural environments of formation of Fe oxides, PO₄ is efficiently scavenged from solution by the Fe oxide sediments.

Figure 5.

Bioavailability of PO₄ from Red Seamount hydrothermal Fe oxide deposits. Growth curve measuring changes in optical density of microbial cultures measured as absorbance at 600 nM, during growth of Acinetobacter ADP1 on minimal salts medium M9 with Fe-PO₄ sediment as a sole P source at room temperature (≈22°C). Fe-PO₄ sediments were rinsed with sterile medium before growth experiments to remove easily desorbed PO₄. No PO₄ was released to the rinse buffer during this procedure.

δ¹⁸O values of Fe-PO₄ samples from the inactive fossil deposits at Green Seamount average 25–26‰ and reflect cold temperatures near that of the ambient seawater (Table 5). The absence of recorded warm temperatures in these deposits is also consistent with their mode of origin as secondary products of the oxidation of hydrothermal iron sulfides at ambient seawater temperatures.

Iron oxide deposits from both Red and Green Seamount also show morphological evidence of microbial activity (29, 33). The iron oxides are composed, almost entirely in some cases, of twisted filaments, stalks, and hollow sheaths that resemble the characteristic remains of the iron-oxidizing bacteria Gallionella ferruginea and Leptothrix known from neutral pH freshwater systems (29). Similar microbial structures have been observed in iron oxide deposits from Loihi Seamount (30). High-resolution structural analyses of iron oxyhydroxides produced by bacteria of the Gallionella and Leptothrix genus indicate that these iron biominerals have distinct structural orientations (34). Recent discoveries, however, suggest that novel iron-oxidizing bacteria that deposit iron oxides not in the morphologically distinct stalks and sheaths of Gallionella and Leptothrix, but as amorphous iron oxide coatings, may be responsible for the majority of microbial iron oxidation in near-neutral pH environments such as seawater (35). Despite strong morphological evidence for the presence of Fe-oxidizing bacteria in some hydrothermal sediments, the importance of microbial activity in formation of hydrothermal iron oxides remains the subject of debate. δ¹⁸O values of PO₄ associated with the microbial iron oxides from Red Seamount provide independent and strong geochemical evidence of metabolic activity within these types of sediments. The association of this PO₄ with iron oxides of probable microbial origin makes Fe-oxidizing bacteria the prime candidate for metabolism of the PO₄. The exact location and identity of microbes responsible for PO₄ metabolism in this system is currently under investigation.

Implications for the Search for Extraterrestrial Life

Mounting evidence for the possible existence of liquid water in subsurface sediments on Mars, beneath the ice-covered surface of Europa, and most recently, Ganymede and Callisto (36), make application of the δ¹⁸O_P biomarker tenable. Systems most likely to be targeted in the search for evidence of extraterrestrial life are those that show evidence of liquid water today (e.g., groundwater seeps on Mars or beneath ice on Europa). In the absence of water for direct comparison of dissolved PO₄ δ¹⁸O_P, certain sediment and mineral phases may be used for δ¹⁸O_P measurement. Specifically, analyses could be made of sediment-associated PO₄ (Fe oxide-bound P) and authigenic phosphate minerals formed by precipitation of dissolved PO₄ from low-temperature fluids, especially occurrences of such phases in sediments and formations interpreted as lacustrine, fluvial, marine, or hydrothermal in origin. Any phosphate that has been processed by biological activity should have an anomalous δ¹⁸O_P value when compared with δ¹⁸O_P of potential PO₄ sources, such as basaltic Martian bedrock.

On Earth, the largest occurrence of biogenic and biologically processed PO₄ is in marine systems (e.g., phosphorite deposits, marine sediments, and dissolved oceanic P). Compared with high-temperature basaltic and igneous sources of PO₄ (5.3‰ to 8‰; refs. 13 and 37), the δ¹⁸O_P of biogenic marine phosphate has a distinctly heavy signature (≈20‰–25‰; refs. 13 and 17) because of the constant, relatively heavy δ¹⁸O value of seawater coupled with the large oxygen isotope fractionations between PO₄ and water at low, near-surface temperatures (Fig. 3). Markel (13) identified biological cycling of phosphate in lake sediments based solely on varying δ¹⁸O_P values of detrital and authigenic apatite mineral phases in the sediments, without requiring the δ¹⁸O value of the water. Anomalous δ¹⁸O_P values in soils have also been attributed to biological activity (37, 38). Thus, δ¹⁸O_P variations and anomalies in sedimentary and authigenic phosphate phases formed at low temperature are also indicative of biological activity. The relatively high concentrations of P measured in Mars surface sediments by Mars Pathfinder (4),^¶ if present as PO₄, would easily allow δ¹⁸O_P analysis.

Conclusions

It has been suggested that life may exist elsewhere in our solar system, in near-surface groundwater on Mars (2) or beneath the ice-covered surface of Europa (39). It is also possible that life evolved 3.5 billion years ago on Mars when hydrothermal systems are believed to have been active (26, 40). Our application of δ¹⁸O_p as a biomarker demonstrates that δ¹⁸O_p values can successfully detect enzymatic activity and microbial metabolism in terrestrial systems. These systems serve as important analogues for extraterrestrial environments that are targeted in the search for life because of the possible presence of liquid water. The δ¹⁸O_p biomarker should provide strong evidence for the presence of life.