The Challenging Detection of Nucleobases from Pre-accretional Astrophysical Ice Analogs

A gas mixture of H₂O:¹³CH₃OH:NH₃ (2:1:1) was first deposited on an inert MgF₂ window at 77 K, forming an astrophysical ice analog. 77 K simulates the position of an icy grain on the edge of a protoplanetary disk where it can receive a sufficient dose of Lyman α photons (Ciesla & Sandford 2012). Irradiation of deposited gas/ice was performed simultaneously during 72 hr at 77 K using microwave plasma with a constant molecular hydrogen flow providing vacuum ultraviolet photons (at Lyman α (121 nm) with a flux of 2E14 photons cm⁻² s⁻¹). After 72 hr of deposition and irradiation, the UV dose in the experiments presented here was 5.2E19 photons cm⁻². This roughly corresponds to a grain submitted to 1.7E6 photons cm⁻² s⁻¹ (Danger et al. 2013) during a life time of the Solar nebula of 1E6 yr (Ciesla & Sandford 2012). The photoprocessed ice was then slowly warmed up until 300 K with 0.1 K per minute. Roughly 100 μg of residue was obtained and kept under argon atmosphere in a stainless steel vessel, to minimize oxidation prior to analysis (de Marcellus et al. 2015). The residue was then solubilized in 150 μL pure methanol and directly analyzed (Danger et al. 2013).

Analyses were performed with two quaternary Accela LC pumps (pump 1: Accela 600 pump; pump 2: Accela 1250 pump) working together and interfaced with a Q-Exactive (Hybrid quadrupole-OrbitrapTM) mass spectrometer equipped with an HESI (Heated Electrospray ionization) source (Thermo Fisher Scientific, Waltham, MA, USA). Peak integration and MS spectra acquisition were performed with Thermo XcaliburTM Qualitative Browser (Thermo Xcalibur 2.2 SP1.48, Thermo Fisher Scientific Inc.). A mass tolerance of 10 ppm was applied for the extraction of targeted ions. Single Ion Monitoring (SIM) chromatograms were traced with Origin (Origin 9.0.0, OriginLab Corporation, Northampton, USA). In order to certify the identification of the targeted compounds, a stock solution of ¹²C nucleobases (cytosine, uracil, thymine, adenine, guanine, lyophilized powder, and Sigma-Aldrich) all at 1E−4 mol l⁻¹ was prepared in methanol.

Triple quadrupole MS analyses were performed with a Shimadzu Nexera X2 UHPLC system coupled to an MS 8050 triple quadrupole mass spectrometer (TQ MS) equipped with an ESI (Electrospray ionization) source (Shimadzu, Marlborough, MA, USA). In order to certify the identification of the targeted compounds, a standard solution containing ¹²C-nucleobases (adenine, cytosine, guanine, thymine, uracil, lyophilized powder, and Sigma-Aldrich) was prepared at 1E−4 mol l⁻¹ in methanol. It was diluted in order to prepare a calibration curve ranging from 5E−8 mol l⁻¹ to 5E−7 mol l⁻¹ (five point-concentrations, each injected in triplicate). Limits of detection (LODs) were calculated from the equation of a linear regression, as three times standard deviation of noise signal/slope ratio. The determined LODs were 1.63E−8 mol l⁻¹ for adenine, 3.34E−9 mol l⁻¹ for guanine, 1.37E−8 mol l⁻¹ for thymine, 1.45E−7 mol l⁻¹ for cytosine, and 1.27E−9 mol l⁻¹ for uracil. Linearity, sensitivity, and reproducibility were ensured.

UPLC-Orbitrap MS was first used to target for the presence of nucleobases similarly as performed by Oba et al. (2019). Thanks to high resolution and high mass accuracy analysis, exact masses of ¹³C nucleobases in residue samples were targeted at the same retention time of the ¹²C nucleobase standards. With this technique three pyrimidine-based nucleobases (cytosine, uracil, and thymine) match in retention time and exact m/z values to their chemical standards. Purine-based nucleobases were not observed. It could be explained by the fact that CO was only formed by photoprocessing of CH₃OH in the here studied ices and is therefore present in lower amounts than in the initial ice studied by Oba et al. (2019). However, UPLC-Orbitrap MS analyses also show that numerous isomers are present (Figure 3). Therefore, providing exact masses at the same retention times as chemical standards could not be a fully satisfactory method to unambiguously reveal the presence of nucleobases.

Zoom In Zoom Out Reset image size

Consequently, UPLC-selected reaction monitoring triple quadrupole mass spectrometry (UPLC-SRM-TQ MS) analysis was performed. In contrast to UPLC-Orbitrap MS, this method specifically uses a fingerprint of a fragmentation pattern of a given molecule. It targets a compound by monitoring simultaneously intact molecules (named the precursor ion) and a particular fragment ion. The pair of a precursor and fragment ion is named "transition." For instance, the fragmentation of the cytosine precursor ion ([cytosine+H]⁺, m/z = 112) resulted in a fragment ion with a m/z = 95 due to a loss of a NH₃ group (m/z = 17), giving evidence for a primary amine group (transition m/z = 112.25 $\to$ m/z = 95.20). Insights into molecular structures of probed analyte molecules are given. Thus, UPLC-SRM-TQ MS enables better discrimination between isomers compared to the UPLC-Orbitrap MS analysis. Following this strategy, we targeted specific transitions of all five canonical nucleobases (adenine, cytosine, guanine, thymine, and uracil) to give evidence for their presence in refractory residues (Figure 4).

Zoom In Zoom Out Reset image size

Only cytosine has been positively identified by matching both retention time and selective SRM molecular transitions (fragmentations) with its ¹²C chemical standard (Nelson & McCloskey 1994; Lu et al. 2006). A limit of detection (LOD) of 1.45E−7 mol l⁻¹ for cytosine was determined, indicating its minimal concentration within the analyzed residue samples.

Based on these findings, we emphasize that high resolution and high accuracy UPLC-Orbitrap MS alone, as performed by Oba et al. (2019), is not sufficient to unambiguously identify specific compounds (e.g., nucleobases) in such complex mixtures including a large diversity of isomers. It is thus absolutely necessary to selectively target respective analytes of interest to identify their presence (e.g., via SRM-LC-MS strategies).

The presence of cytosine in pre-accretional residue analogs suggests that this nucleobase has been formed in astrophysical environments via radical chemistry. This hypothesis is in agreement with the detection of its putative precursor building blocks (Sadr-Arani et al. 2015) formed by UV processing of astrophysical ice analogs, e.g., HNCO (Fedoseev et al. 2016), OCN⁻ in the presence of NH₃, Figure 5 (Hudson et al. 2001), HNC (Quinto-Hernandez et al. 2010), or NH₂ radicals (Borget et al. 2015; Fedoseev et al. 2016). In contrast, the mechanism of formation of the other four nucleobases remains unclear. They are either formed by radical chemistry as well but resulting in lower yields, or via an alternative mechanism. Thus, CO in ices (Oba et al. 2019) may be significant reactants in the formation of nucleobases. Alternatively, nucleobases might be formed more efficiently in aqueous environments in parent bodies of meteorites (Robertson & Miller 1995; Saladino et al. 2001) which could explain the higher abundance of guanine observed in meteorites (Callahan et al. 2011).

Zoom In Zoom Out Reset image size

The evolution of extraterrestrial organic matter, starting from dense molecular clouds toward planetary systems can be experimentally simulated in the laboratory (Öberg 2016). Collapsing clouds lead to the progressive formation and evolution of the solar nebula that evolves toward the solar system formation (and more generally to any planetary/exoplanetary system). During this evolution, icy grains constitute the initial solid reservoir in molecular clouds, as dictated by cosmic abundances and CHNOPS-based organic chemistry (d'Hendecourt 2011). This starting solid molecular pool, primarily made out of hydrides (Oort & Van de Hulst et al. 1946), will evolve toward an increasing molecular diversity and molecular complexity (Schmitt-Kopplin et al. 2010; Danger et al. 2013, 2016; Fresneau et al. 2017; Ruf et al. 2018), driven by energetic processes surrounding icy grains. If one wishes to better formalize this hypothesis, we might consider the holistic model of Ciesla & Sandford (2012). This model considers turbulent radial and convective dust grain transport, in and out of plane, that brings icy grains into outer regions of the solar nebula. These outer grains are then directly exposed to UV light, either from the central star or, more importantly, from UV light of surrounding stars (Gounelle & Meynet 2012).

Such energetic processes enrich the organic chemistry that has already been present on individual grains and has got furthermore incorporated into asteroids or comets, the "leftovers" of incomplete accretion processes (Öberg 2016). These interplanetary bodies might have played a decisive role in "shaping" the ecospheres of telluric planets, as this might have been the case for Earth (Catling & Kasting 2017). In these small bodies, organic matter can endure secondary processing, mainly in asteroids (Gounelle 2011). Meteoritic analyses give insight on these secondary processing, such as hydrothermalism or metamorphism, that can lead to a reprocessing of the accreted organic content (Vinogradoff et al. 2018). In comets, much weaker secondary processing is presumed to occur. Comets represent reservoirs of the most primitive organic matter (Cottin et al. 1999; Brearley 2006). Asteroids or comets are then able to deliver their organic content onto the surface of a telluric planet, particularly during the period of the Late Heavy Bombardment where interplanetary dust particles/micrometeorites, asteroids, and comets could have been the main drivers of organic matter at the surface of the early Earth (Chyba et al. 1990; Whittet 1997).

All these evolutionary processes can be systemically simulated by laboratory experiments, as presented in this work. These experiments have shown that individual irradiated icy grains lead to a large molecular diversity, prior to their incorporation in minor solar system bodies (Danger et al. 2013, 2016; Fresneau et al. 2017). Specific molecules have been detected that might have played a key role in emergence of life processes, e.g., amino acids (Muñoz Caro et al. 2002) or sugars (Meinert et al. 2016). Here, we demonstrate that at least one nucleobase (cytosine) is formed during the same ice processing, without any aqueous alteration effects (e.g., catalytic reactions of minerals).

Consequently, all major prebiotic building blocks (amino acids, sugars, and nucleobases) might have already been present in the solid phase (icy grains) of forming protoplanetary disks, before accreting toward planetary systems. So, is there a direct pathway in chemical evolution from astrochemistry toward prebiotic chemistry? While in asteroids secondary alteration could occur, comets could have delivered their primitive, unaltered organic content to the surface of telluric planets. Thus, comets could be the direct link between organic matter which has been formed during solar nebula evolution and the surface of telluric planets. Once on a surface of telluric planets, these exogenous prebiotic building blocks could have acted as an organic source toward the emergence of evolved prebiotic systems, as might have been the case in Earth's early environment.

Before injection, samples were stored at 4 °C using a Stack cooler CW (CTC Analytics AG, Zwingen, Switzerland). MS functions and LC solvent gradients were controlled by the Xcalibur data system (Thermo Fisher Scientific). Acetonitrile with $0.1 \%$ formic acid was used as buffer A and water with $0.1 \%$ formic acid as buffer B. Instrument calibration in positive mode was done every day via the direct infusion of positive ion calibration solutions. Compounds elution was performed via a 1D-LC configuration composed of three serially coupled columns (Eddhif et al. 2018). It combined a porous graphitic carbon Hypercarb (50 mm × 2.1 mm, 3 μm, 250 Å; Thermo Fisher Scientific), a Pentafluorophenyl reverse-phase Kinetex F5 (100 mm × 2.1 mm, 1.7 μm, 100 Å, Phenomenex) and a C18 reverse-phase Hypersil Gold Q (50 mm × 1 mm, 1.9 μm, 175 Å, Thermo Fisher Scientific). The elution was performed at a constant flow of 100 μl minute⁻¹ within the series. The elution gradient started with 5% buffer A and 95% buffer B during 3 minutes. It continued with a first linear gradient to reach 35% buffer A and 65% buffer B in 12 minutes. A second linear gradient was employed to reach 60% buffer A in 5 minutes. A third linear gradient was performed to reach 100% of buffer A in 1 minute. The column was rinsed with 100% buffer A during 2 minutes and then reconditioned for 2 minutes with 5% buffer A and 95% buffer B. Target Selected Ion Monitoring data dependent-MS/MS (t-SIM-ddMS/MS) was used to detect nucleobases in the sample. For chemical standards, the precursor ions selected for cytosine, uracil, thymine, adenine, and guanine were, respectively, [M+H]⁺ m/z 112.05054, [M+H]⁺ m/z 113.03512, [M+H]⁺ m/z 127.05669, [M+H]⁺ m/z 136.06180, and [M+H]⁺ m/z 152.05670. Mass detection was performed with an electrospray voltage at 4.0 kV. Capillary and heater temperatures were set to 280 °C and 300 °C. The relative flow rate of the sweep gas (nitrogen) which aids in solvent declustering and adduct reduction was set to 0 (flow in arbitrary units as defined by Thermo Fisher Scientific). To help nebulize sample solution into a fine mist in the ESI nozzle, the relative flow rate of the sheath gas (nitrogen) was set to 40 (flow in arbitrary units as defined by Thermo Fisher Scientific). Finally, the relative flow rate of the auxiliary gas (nitrogen) which assists the sheath gas in dispersing and/or evaporating sample solution was set at 30 (flow in arbitrary units as defined by Thermo Fisher Scientific). Targeted MS parameters were optimized as follows: resolution of 17,500 for precursor ions and 17,500 for fragment ions, AGC target of 5.105 (precursor ion) and 2.104 (fragment ions), microscan 1, max IT of 100 ms (precursor ion) and 200 ms (fragment ions), MSX count 4, loop count 3, and isolation window of 2.0 m/z. The normalized collision energy was set to 35%.

Before injection, samples were stored at 4 °C using a sample cooler (Shimadzu). LC solvent gradients and MS functions were controlled by the LabSolution data system (LabSolution 5.89, Shimadzu). Water with $0.1 \%$ formic acid was used as buffer A and acetonitrile with $0.1 \%$ formic acid as buffer B. An LC configuration composed of three serially coupled columns was used (Eddhif et al. 2018). It combined a porous graphitic carbon Hypercarb (50 mm × 2.1 mm, 3 μm, 250 Å; Thermo Fisher Scientific), a Pentafluorophenyl reverse-phase Kinetex F5 (100 mm × 2.1 mm, 1.7 μm, 100 Å, Phenomenex) and a C18 reverse-phase Hypersil Gold Q (50 mm × 1 mm, 1.9 μm, 175 Å, Thermo Fisher Scientific). The LC oven temperature was set at 40 °C. The elution was performed at a constant flow of 100 μl minute⁻¹ within the series. The elution gradient started with 5% buffer B and 95% buffer A during 3 minutes. It continued with a first linear gradient to reach 35% buffer B and 65% buffer A in 12 minutes. A second linear gradient was employed to reach 60% buffer B in 5 minutes. A third linear gradient was performed to reach 100% of buffer B in 1 minute. The column was rinsed with 100% buffer B during 2 minutes and then reconditioned for 6 minutes with 5% buffer B and 95% buffer A. The TQ MS instrument operated in SRM mode to detect nucleobases in the sample. The parameters of the mass spectrometer were set as follows: capillary voltage of 4 kV; interface temperature of 400 °C; heat block temperature at 500 °C; and desolvation line temperature at 250 °C. Nebulizing gas flow, heating gas flow, and drying gas flow were set to 180 l hr⁻¹, 600 l hr⁻¹, and 600 l hr⁻¹, respectively. Respective SRM transitions (for ¹²C compounds), dwell times, and collision energies are summarized as follows: cytosine, 112.25 $\to$ 95.20 for 12, 5 ms, 22 V; uracil, 113.25 $\to$ 96.15, 10 ms, 19 V; thymine, 127.25 $\to$ 110.15, 1 ms, 21 V; adenine, 136.25 $\to$ 119.20, 10 ms, 25 V; and guanine, 152.25 $\to$ 135.15, 10 ms, 21 V (Nelson & McCloskey 1994; Lu et al. 2006).