Introduction

A central question in understanding biogenesis is the nature of the process that provided ammonia for the sustained synthesis of the molecules essential to the origin of life. To develop a model of nitrogen fixation in which ammonia is simultaneously synthesized and incorporated into organics, we investigated the sugar-driven reductive synthesis of ammonia from nitrite in the absence and presence of small amounts of aqueous Fe+3 catalyst. The sugar-driven synthesis of ammonia from nitrite is considered important to abiogenesis because it (1) provides a way to generate ammonia at sites of sugar-based origin-of-life processes, and (2) eliminates the need for a planet-wide source of photochemically unstable ammonia for abiogenesis (Cleaves 2008, refer. therein). Our earlier studies showed that sugars react with ammonia and amines under prebiotic conditions yielding molecules with catalytic, energy-transfer, and potential replication properties, as well as the ability to form cell-like structures (Weber 2001, 2002, 2004, 2005, 2007, 2008, refer. therein). Nitrite is considered a plausible prebiotic precursor of ammonia because atmospheric shock heating (lightning, coronae, meteors) of N2 is known to generate NO that can be photochemically converted to HNO that transforms to nitrite and nitrate in water [HNO ➔ NxOx ➔ NO2 −1 + NO3 −1] (Summers and Khare, refer. therein 2007). Model prebiotic studies have also shown ammonia synthesis from nitrite using Fe+2 and FeS as the reductant (Summers and Chang 1993; Summers 2005).

Materials and Methods

Materials

dl-Glyceraldehyde dimer (95%), dihydroxyacetone (97%), formaldehyde, glycolaldehyde, pyruvaldehyde, ascorbic acid, ammonium chloride (99.998%), glacial acetic acid (99.8%), sodium hydroxide (99.99%) and sodium citrate trisodium dihydrate were obtained from Aldrich; d-ribose, d-glucose, sodium nitrite, sodium nitrate, acetonitrile (HPLC grade) from Sigma; d-erythrose from Fluka; ferric chloride 6-hydrate from J. T. Baker; o-phthaldialdehyde from Molecular Probes, and 1 ml vacules from VWR. Previously we showed that the Aldrich dl-glyceraldehyde contained about 64% glyceraldehyde and 36% dihydroxyacetone (Weber 2007).

Reaction Method

Sugar-nitrite reactions were carried out in heat-sterilized 1 ml vacules using the concentrations and conditions described in the figures. The vacules containing the reaction solutions were sealed in vacuo after being deaerated in a frozen state by cycling four times between a vacuum and nitrogen gas. The reactions were carried out at 65°C in a Labnet (Model 611D) incubator, and then stored at −80°C until analyzed.

Ammonia Derivatization for HPLC Analysis

Ammonia was measured using an ammonia-selective HPLC method based on o-phthaldialdehyde (OPA) derivatization (Goyal et al. 1988). Ammonia derivatization involved adding a 10 µl aliquot of each reaction solution to 290 µl of OPA reagent (5 mg of o-phthaldialdehyde, 50 µl of 1.0 M mercaptoethanol, and 3.7 ml of 0.10 M sodium phosphate buffer (pH 6.8) in a 1 ml reaction vial. The derivatization vials were sealed with Teflon caps, and then heated 5 min at 65°C with magnetic mixing in a Reacti-Therm heating block with 1 ml water in the wells in the aluminum block to improve heat transfer to the reaction vials. A 20 µl aliquot of the OPA-derivatization product was analyzed by HPLC using a Beckman Model 126 HPLC system equipped with a JASCO fluorescence detector. The OPA-ammonia product was measured by its fluorescence (excitation 410 nm, emission 470 nm). HPLC analysis used an Alltech Alltima-C18 column (5 µ, 150 × 3.2 mm) eluted with 36% acetonitrile in 30 mM sodium acetate (pH 5.4) pumped at 0.7 ml/min. Elution time of the ammonia-OPA adduct was 5.2 min.

Results and Discussion

Concentration Dependence of Sugar-Driven Ammonia Synthesis from Nitrite

We discovered that sugars, and some sugar precursors and products, drive the reductive synthesis of ammonia from nitrite under mild aqueous conditions, a synthetic process that is catalyzed by µM amounts of Fe+3 cation. Figure 1 shows the time course of glyceraldehyde-driven synthesis of free ammonia measured at three different glyceraldehyde concentrations (free NH3 is defined in the Fig. 1 caption). Ammonia was not formed in control reactions that lacked glyceraldehyde or nitrite, or had nitrate substituted for nitrite. The figure shows that the free ammonia yield based on nitrite at 21 days reached 4.8%, 5.8%, and 6.2% using 3.2 mM, 5.5 mM, and 16.5 mM glyceraldehyde, respectively. These are not final yields because there is only a small decrease in the rate of ammonia synthesis in the 10–21 day time interval. Surprisingly, these measurements show that increasing the glyceraldehyde concentration 5-fold from 3.2 mM to 16.5 mM caused only a 24% increase in the free ammonia yield at 21 days. This unusual result, together with the knowledge that glyceraldehyde is known to react with ammonia to give nitrogenous organics (Weber 1998, 2008), prompted us to run a control reaction of 16.5 mM glyceraldehyde with 0.25 mM ammonia (a concentration equal to the 5% NH3 yield in Fig. 1). This control reaction showed that 16.5 mM glyceraldehyde reacted with ammonia causing a 34% (8 days) and 48% (16 days) decrease in the free ammonia concentration. This result establishes that the free ammonia values plotted in Fig. 1 are measurements of the difference between glyceraldehyde-driven ammonia formation and the concurrent reaction of part of the ammonia product with glyceraldehyde, a reaction known to incorporate ammonia into nitrogenous organic products (see reaction scheme-1, Weber 1998, 2008). A similar control reaction showed that dihydroxyacetone, unlike glyceraldehyde, reacted very slowly with ammonia resulting in only a 2% and 4% decrease in detectable ammonia at 8 and 16 days, respectively.

Fig. 1
figure 1

Time course of glyceraldehyde-driven free ammonia synthesis from 5.0 mM nitrite at 65°C in 75 mM sodium acetate (pH 5.5). Free ammonia is the ammonia product that reacts with the o-phthaldialdehyde derivatizing agent to give the ammonia-OPA fluorescent adduct measured by HPLC

Full size image
(1)

Reductant Specificity and Ferric Ion Catalysis of Ammonia Synthesis

Figure 2 below shows measurements of ammonia synthesis from nitrite driven by various sugars, formose sugar precursors, and sugar reaction products. The figure shows that the reductive synthesis of ammonia from nitrite was driven by glyceraldehyde, dihydroxyacetone, erythrose, ribose, glycolaldehyde, pyruvaldehyde, and glyoxal. However, under the same reaction conditions glucose, 2-deoxyribose, formaldehyde, and glyoxylate did not produce measurable amounts of ammonia from nitrite. Also, the ammonia yield at 8 days declined progressively going from smaller to larger sugars: triose > tetrose > pentose > hexose = 0. This ordering suggests that the rate of ammonia reduction depends on the availability of the acyclic form of the sugar reductant that decreases in the same order. The lower free ammonia yield of glyceraldehyde compared to dihydroxyacetone is probably partly due to glyceraldehyde’s faster rate of reaction with ammonia (described earlier) that incorporates more of the free ammonia into organic products. Figure 2 also shows that all the reactions that yielded ammonia were catalyzed by addition of 50 µM ferric chloride. Reactions that did not produce ammonia were not affected by the addition of ferric chloride. In the case of glyceraldehyde, a ferric cation concentration as low as 5 µM noticeably increased ammonia synthesis.

Fig. 2
figure 2

Free ammonia synthesis driven by sugars, their precursors and products. Reactions were carried out at 65°C for 8 days using 16.5 mM reductant, 5.0 mM nitrite in 75 mM sodium acetate (pH 5.5)

Full size image

Proposed Reaction Pathways

Figure 3 shows proposed reaction pathways for the sugar-driven reduction of nitrite to ammonia. These pathways are consistent with (1) the known sugar transformations of isomerization, enolization, and β-dehydration (Speck 1958; Feather and Harris 1973), (2) the results here showing that only molecules containing an α-hydroxyaldehyde, α-hydroxyketone, or α-ketoaldehyde group capable of forming enediolate and aldehydrol anion intermediates are able to reduce nitrite to ammonia, (3) the established involvement of the anionic form of aldehyde hydrates (anhydrols) in hydride transfer (Fratzke and Reilly 1986; Watt 1988), and (4) studies showing that aldonic acids are the major product of sugar oxidation by iron(III) oxide hydroxide and several other oxidants (Green 1980; Weber 1992). As shown, the pathways begin by the reversible isomerization of aldoses yielding their respective ketoses, followed by β-dehydration to give their respective 3-deoxy-aldos-2-uloses (e.g. pyruvaldehyde from trioses). Alternatively, by Pathway-A1 the aldoses and ketoses can enolize and ionize producing the α-hydroxyaldehyde enediolate intermediate (structure I) that either alone, or complexed to Fe+3 as the monoanion (structure II) or dianion (not depicted) drives the reduction of nitrite to ammonia. Since glycolaldehyde cannot undergo β-dehydration, its transformation is limited to Pathway-A1. As shown in Pathway-A2, the 3-deoxy-aldos-2-ulose products of sugar β-dehydration can hydrate and ionize to give α-ketoaldehyde aldehydrol anion intermediate (structure III) that either alone, or complexed to Fe+3 as the monoanion (structure IV) or C2-hydrate dianion (not depicted) drives the reduction of nitrite to ammonia. It is possible that Fe+2 instead of Fe+3 is involved in catalysis, because glyceraldehyde has been shown to reduce Fe+3 in iron(III) oxide hydroxide to Fe+2 under similar conditions (Weber 1992). However, it is unlikely that Fe+2 directly reduces nitrite to ammonia in the pH 4.5–5.5 reactions described here, because other studies have shown that nitrite reduction to ammonia by Fe+2 requires a pH ≥ 7 (Summers and Chang 1993). Figure 3b shows intermediates possibly involved in nitrite’s reduction to ammonia, a process requiring three successive hydride transfers. As shown, nitrite reduction oxidizes sugars to their α-hydroxy acids (e.g. glyceric acid from glyceraldehyde), and α-ketoaldehydes to their 3-α-keto acids (e.g. pyruvate from pyruvaldehyde). This oxidation process is probably limited to one hydride transfer per sugar molecule, because the α-hydroxy acid products lack a carbonyl group needed for hydride transfer, and the 3-deoxy-α-keto acid products are missing the 3-hydroxyl group required for hydride transfer. However since organic intermediates and products of this reduction have not been identified, it is also possible that the oxidation of the sugar α-hydroxycarbonyl group (–CHOH-CO-) produces an α-ketoaldehyde group (–CO-CHO), as described for the alkaline oxidation of sugars by Fe+3 (Glor 1968). Unlike the oxidation process that yields hydroxy acids and keto acids, this oxidation process would allow multiple hydride transfers from a single sugar molecule, because it yields α-ketoaldehyde groups capable of further hydride transfer.

Fig. 3
figure 3

Proposed pathways for a the formation of the α-hydroxyaldehyde enediolate and and α-ketoaldehyde aldehydrol anion intermediates involved in the reduction of nitrite to ammonia, and b the reduction of nitrite to ammonia showing possible intermediates: NO- or N2O, and NH3OH+. In structures (I–IV) the transferred hydride is indicated by an oversized, bolded hydrogen symbol

Full size image

Prebiotic Relevance

The ability of sugars to produce ammonia from nitrite described here, together with their high self-transformation energy (Weber 2000), give sugar-based origins models a high degree of synthetic self-sufficiency. In fact, since small sugars and their precursors are known to react with ammonia to produce a variety of nitrogenous products (Weber 2001, 2002, 2004, 2005, 2007, 2008, refer. therein), a sugar-based origins model yielding nitrogenous multicarbon products requires only a prebiotic source of formaldehyde, glycolaldehyde, and nitrite. Within the context of current knowledge, formaldehyde, glycolaldehyde, and nitrite could have been generated in the Earth’s primitive atmosphere, and delivered to the surface by rainout (Summers and Khare 2007; Cleaves 2008, ref therein) where they reacted to give sugars, ammonia, and ultimately nitrogenous organic products such as cell-like structures, biometabolites, catalytic molecules, energy molecules (thioesters), and plausible alternative nucleic acid coding bases (Weber 2001, 2002, 2004, 2005, 2007, 2008, ref. therein). Furthermore, since sugar-driven synthesis is a “one pot”, spontaneous, catalyzable, aqueous process, it does not require coupling to any additional source of chemical energy, and is amenable to catalysis within protocellular organic structures. Because both the sugar-based prebiotic synthesis and the modern biosynthesis of small molecules involve and are driven by the redox disproportionative transformation of sugars (Weber 1997, 2000, 2002), prebiotic sugar-based synthetic processes also have the potential to evolve directly into modern sugar-driven biosynthesis without violating the principle of evolutionary continuity (that is, without replacement of the core synthetic process).