1. Introduction

Francis Crick believed that the problem of the origin of life on Earth or other planets is one of the great unsolved problems of Biology and in general of Science. To study the problem of the origin we have to answer to the basic question raised in 1944 by Schrődinger (Schrődinger 1944): “What is life”, that remains until today without a definitive answer, because it is not scientifically evident how set a boundary between living and not living systems. However, the analysis of modern living systems leads to the observation that they are all based on a common organization: a mechanism to store and transmit the genetic information, a metabolism that provides energy and a boundary between inside and outside of the cell. Moreover, all living systems are dissipative structures that require a flux of free energy from the outside. These are the fundamental aspects of life, and although other features are necessary, the presence of genetic material is absolutely essential. The appearance in an ancestral era of a polymer able to replicate and evolve, from the Darwinian point of view, marks the beginning of life (Lazcano 2010). The 'RNA World' theory assumes that RNA-like molecules were not only the carrier of genetic information but also the enzymes that performed catalytic activity. The hypothesis, originally proposed by Gilbert (Gilbert 1986) and Orgel and Crick (Orgel and Crick 1993), has been strengthened by the discovery of catalytic RNA molecules, the ribozymes (Guerrier-Takada et al. 1983; Kruger et al. 1982), and by the fundamental role played by RNA in many biological processes, particularly in ribosome structure and function (Moore and Steitz 2002; Steitz and Moore 2003). More recently, it has been proved - for the first time in vitro - the possibility of self-replication and exponential growth of an RNA system without any catalytic support from protein, opening the way for the solution of the “chicken and egg” paradox (Lincoln and Joyce 2009) of genetic information and catalysis, in pre-cellular systems.

In nucleic acids catalytic and genetic functions are based on the specificity of base pairing of nucleobases. The complementarities of adenine-uracil (thymine) and cytosine-guanine are one of the chemical process on which life (as we known) is based. It is a “simple and efficient” mechanism for storage and propagation of genetic information by mean of template replication, so ensuring the persistence of the genetic program. At the same time it provides the basic mechanism for RNA polymers to fold in the specific 3d structure required for catalysis.

From this point of view, it is clear that the presence and availability of adenine, cytosine, uracil and guanine is a fundamental prerequisite for any RNA world like scenario. However, it was pointed out by different authors that synthesis and concentration of precursor of nucleic acids would be not an easy task in prebiotic condition. In particular, Shapiro (Shapiro 1999) has outlined that synthesis and stability of cytosine appears to be a major problem in prebiotic RNA evolution that compromises the entire foundation of RNA World theory.

In recent year, it has been proposed that mineral matrices could have represented the possible cradle were the replicators, RNA- like molecules, could have been originated and evolved (Gallori et al. 2006).

The deep relations between mineral matrices and evolving replicators, in pre -cellular systems, leads to the hypothesis of an “RNA-mineral-World” (Biondi et al. 2007a; Biondi et al. 2007b). This formulation extends the original idea of Bernal (Bernal 1951), who first recognized that clay minerals could have played a pivotal role in prebiotic chemistry. The RNA-mineral-World hypothesis has received, in recent years, ample support from both theoretical and experimental work. It helped to solve some, thus far inaccessible, problems of the RNA world hypothesis, such as the synthesis and concentration of precursors, the polymerization of the precursors into larger molecules, and their protection against degradation (Baaske et al. 2007; Biondi et al. 2007b; Costanzo et al. 2007; Ertem and Ferris 1996; Ferris et al. 1996; Koonin 2007; Koonin and Martin 2005; Ricardo et al. 2004; Saladino et al. 2004).

In this report we focus our attention on the stability of cytosine adsorbed on montmorillonite. The fate of cytosine is particularly relevant in a prebiotic context because of the hydrolytic deamination to uracil represents one of the major chemical challenges to the RNA world theory. We studied the cytosine-clay system exposed at high temperature for different periods of time. Powdered x-ray diffraction (XRD) and spectroscopic (FTIR-DRIFT) techniques were used to study the processes of interaction and degradation of adsorbed cytosine. We choose FTIR-DRIFT because it is a non invasive technique to study at molecular level the interaction between organic molecules and clay (Pucci et al. 2008). The choice of montmorillonite was dictated by: i) the fact that clay minerals would be common and abundant on the early Earth , ii) their possible role in prebiotic chemistry has been exhaustively investigated (Biondi et al. 2007a; Biondi et al. 2007b; Ertem and Ferris 1996; Ferris et al. 1996; Hanczyc et al. 2003; Ten et al. 2001).

The study of the kinetics of degradation of cytosine in the presence of minerals can lead a better understanding of interplay between mineral matrices and organics in the context of the origin of life. Our main result is that cytosine stability is markedly increased when strongly interacting with a mineral matrix.

2. Methods

2.1. Samples preparation Cytosine adsorption on clay.

A source clay (SWy-2, Crook County, Wyoming, USA) from the Clay Minerals Society was fractionated (<2μm) by sedimentation and made homoionic according to the method of Sposito (1981). To 9 ml of 10mM aqueous solutions of cytosine (Sigma- Aldrich) was added 1 ml of clay suspension (2 mgml-1 montmorillonite) to a final volume of 10 ml and agitated in an end-over- end shaker at 15 rev min-1 for 24 h to reach equilibrium. The final pH 5.5 for each sample was adjusted by adding 1 M HCl. The untreated clay and the clay-cytosine suspensions were centrifuged. The sedimented clays were air-dried at room temperature (298 K), weighed and powdered.

2.2. Infrared spectroscopy

Samples obtained after drying the suspensions were prepared for IR spectroscopy by using the KBr disk method. The powders diluted 1:10 with KBr (Fluka) were finely ground and pressed into a disk at 8 t under vacuum by a hydraulic press.

Diffuse reflectance infrared Fourier transform (DRIFT) spectroscopic measurements were performed by Spectrum GX2000 spectrometer (Perkin-Elmer) using a diffuse reflectance accessory (Perkin-Elmer). Spectra were recorded from 4000 cm-1 to 400 cm-1 with a resolution of 4 cm-1 and corrected for effects of atmospheric CO2 and water vapour. The reflectance spectra expressed as Kubelka-Munk units versus wave number curves are very similar to absorbance spectra and can be evaluated accordingly. The Norton-Beer function was used for apodization.

2.3. X-ray diffraction

X-ray diffrattograms of powdered samples were collected by X’Pert PRO diffractometer (Panalitycal) with Cu Kα radiation (λ = 1.5418), equipped with an Anton Parr HTK 1200N, high temperature oven chamber. The diffractograms of the untreated clay and the room temperature (RT) (52% RH, relative humidity), is 1.25 nm due to the presence of water molecules adsorbed within the monomolecular sheet (Brindley and Brown 1980; Norrish 1954).

3. Results and Discussion

3.1 X-ray diffraction Smectite clays have the properties of cation exchange, intercalation of molecules, and swelling in solvents (Pinnavaia 1983). The c-axis spacing in the Na-montmorillonite, at room temperature (RT) (52% RH, relative humidity), is 1.25 nm due to presence of water molecules adsorbed in monomolecular sheet (Brindley and Brown 1980; Norrish 1954). In Figure 1A untreated montmorillonite after heating at 448 K for 72 h displays a change of basal spacing from 1.24 nm to 0.96 nm. These values corresponded to the hydrated and dehydrated forms of the clay. These results are in agreement to those reported by Brindley and Brown (1980) where the reduced d001 basal spacing rely on the loss of interlayer water resulting in a “collapsed” structure.

In contrast, when cytosine was adsorbed on clay (Figure 1B) the profile of the basal reflection does not change after similar heating treatment. This suggests that: i) the interlayer space is occupied by cytosine with a molecular configuration wherein the ring is parallel to the clay layer giving a 1.25 nm spacing (see also Lailach et al. 1968; Serratosa 1966); ii) cytosine remains stably intercalated in the interlayer of montmorillonite at high temperature (448 K for 72h).

3.2. Infrared Spectroscopy

The DRIFT-FTIR spectra of solid anhydrous cytosine (a), powdered montmorillonite (b), and cytosine adsorbed onto Na-montmorillonite (c), at RT in the 1600/1800 cm-1 range, are reported in Figure 2. The bands lying in this range are important for their sensitivity of molecular interactions between clays and nucleobases, as as pyrimidines absorb strongly in this region.

The cytosine solid spectrum (a) shows two peaks at 1661 cm-1 and 1616 cm-1 assigned to pyrimidine aromatic stretching bands (1vC₅C₆,νC₂O,νC₆N1,νC₂N₃, Figure 2, Table 1) and to a δNH bending coupling among stretching vibrations νC₅C₆,νC₂0, respectively (Figure 2, Table 1) (Florian et al. 1996). The spectrum of Na-montmorillonite (b) shows in this region a shoulder at 1636 cm-1 assigned to the -OH bending of water coordinating the clay matrix (Farmer 1974) (Figure 2, Table 1), and only partially affects overlap with cytosine spectra. The spectrum of cytosine adsorbed onto montmorillonite (c) at pH= 5.5 shows a peak at 1730 cm-1 assigned to protonated cytosine (vC₂=O,δN₃H,vN₁C₂,δN₁H) interacting with Na-montmorillonite. This is probably after the peak at 1661 cm^-1 was shifted to 1730 cm-1 by formation of cytosine-H+ species in good agreement with the data reported for protonated cytosine (Lailach et al. 1968) (Figure 2, Table 1). The spectrum of cytosine adsorbed onto montmorillonite (Figure 2c, Table 1) shows a second peak at 1685 cm-1. This peak is probably due to cytosine molecules weakly interacting with clay. In this case the characteristic peak of cytosine at 1661 cm-1, formerly assigned to pyrimidine aromatic stretching bands, is shifted to a significantly lower frequency (1685 cm-1) due to the complementary H-bonding with clay matrix.

Finally, we observe a weak shoulder at 1623 cm-1. This was assigned to the peak shifted from 1616 cm-1 of free solid cytosine overlapping with δOH bending of water coordinated the clay matrix (Figure 2, Table 1).

These findings suggested that cytosine is taken up in two different ways: 1) a loosely adsorption by hydrogen bond probably mediated by interaction with water molecules with a characteristic peak at 1685; and 2) a strong adsorption by cation exchange, where a protonated cytosine (CH+) substitutes for Na+ in Na-montmorillonite with a characteristic peak 1730. The exchange process of protonated cytosine can be directed (equations 1-2) or mediated by formation of H-montmorillonite. In this case interlayer cations of the clay are partially replaced by protons, due to cation exchange reactions, and the resulting H-montmorillonite may adsorb neutral cytosine molecules by their subsequent protonation (equations 3-4) (Lailach et al. 1968). The most probable protonation sites of cytosine could involve the O(2) and N(3) atoms of the aromatic ring (Florian et al. 1996).

At this stage, the adsorption is dependent upon the pH (equation 1) of the equilibrating solutions, but it was pointed out that the proton concentration in the interlayer may be significantly higher than in the solution, because the clay itself becomes partially an H-clay (dependenting on exchangeable cations) (Lailach et al. 1968).

Figure 3 displays the DRIFT-FTIR spectra of cytosine adsorbed onto clay after heating (T= 448 K) for different exposure times (0, 4, 8, 16, 24 and 72 h). During heating, the adsorption profile of samples changes further with the time exposure: the peak at 1730 cm-1 (attributed to cytosine tightly interacting with clay) was observed in all spectra whereas the peak at 1685 cm-1, previously attributed to loose interaction with cytosine, disappears during the early stage of heating(Figure 3).

After 72 h of treatment, a new peak at 1703 cm^-1 appears (Figure 4) that is likely due to uracil. In fact, uracil adsorbed onto clay exhibits a peak at 1670 cm^-1 and 1710 cm^-1 attributed to νC₅C₆,νC₄O₁₀,δC6H and νC₄O₁₀,νC₄C₅,νC₅C₆ respectively (Ten et al. 2001). These peaks were maintained during the heating at 448 K (data not shown).

Figure 5 shows the variation of the intensities of peaks at 1685 cm^-1 and 1730 cm^-1 as function of time at 448 K. In order to compare the intensities, they are reported as a fraction of their value at t = 0, so that the initial value is equal to 1 by definition. Each point represents the mean of at least 10 independent experiments and the error bar is the standard error. The solid line is the best fitting of the experimental data obtained using exponential decay models. Estimation of fitting parameter and statistical analysis was obtained using software package in Matlab 7.5.0.

Data obtained clearly reveal the high rates of decay of the peak at 1685 cm^-1 compare to the peak at 1730 cm^-1. Moreover, the intensity of the peak at 1685 cm^-1 is almost 0 after 72 hours, whereas the peak at 1730 cm^-1 was reduced by nearly 40 % in the same amount of time.

Data fitting show that the decrement of peak at 1685 cm-1 at 448 K could be described by a single exponential decay function R=e−kt, where R is the fraction of residual intensity relative to the initial value at t = 0 and k = 0.3 h-1 is the decay rates of peak intensity.

The behavior of peak at 1730 cm-1 was more complicated and a better description of the decay process was obtained using a linear combination of two exponential functions:

⁽²⁾

where x is the fraction of population 1 respect population 2. The best estimation of parameters was: x = 0.21, kA = 0.07 h^-1, kB = 0.0003 h^-1.

The simplest interpretation of these findings is to assume that the peak at 1730 cm^-1 is assigned to two different populations of protonated cytosine. A first group (I) (~ 21 % of the total) shows a faster decay rate (k₁ = 0.07 h^-1) at 448 K. A second group (II) (~ 79% of the total) shows a decay of three order of magnitude (k₂ = 0.0003 h^-1) compared to that for cytosine loosely bound to montmorillonite.

The difference in decay of protonated cytosine is not easily interpreted and further analysis should be made to clarify this point. However we could speculate that the fraction of protonated cytosine that is degraded faster (group I) comprises protonated cytosine dissolved within the interstices of the clay, or adsorbed on its external surface. Whereas, we suggest that the reduced degradation and consequently the thermal stability observed in group II (protonated cytosine) is due to strongly interacting molecules with the mineral matrix. The protection could be provided by the intercalation of the molecules in the interlayer of montmorillonite as shown by X-ray spectra.

Results obtained indicated an increased thermal stability of cytosine in presence of montmorillonite particles. However, the increase of two orders of magnitude in thermal stability of cytosine when tightly adsorbed, supports the hypothesis that mineral particles could have provided a suitable niche where chemical evolution might have begun. Because of the change of the physical and chemical properties of the molecule when adsorbed on mineral matrices, the stability of cytosine could have been compatible with the early scenario of chemical evolution, if adsorbed on mineral matrices, a finding that addresses some of the concerns of Shapiro (1999).

Moreover, the differential degradation of loosely and tightly adsorbed cytosine could provide a mechanism for local enrichment of key precursors of RNA molecules. Further investigations are necessary to better quantify the presences of degradation products such as thymine. In a wider perspective the interplay between minerals and organic molecules could represent a fundamental aspect in the context of prebiotic evolution. Here we want to outline that mineral matrices not only could have provided support and the compartmentation necessary for prebiological chemical evolution (Branciamore et al. 2009; Martin and Russell 2003; Martin and Russell 2007), but also the interaction between organic molecules and minerals could modify structural and physico-chemical properties of the matrix itself, thus changing the micro environment where evolution occurs. For example, organic molecules could act to promote the aggregation of small mineral particles to form larger stable aggregates as they do for humic substances (Stevenson 1994). The increased complexity in mineral matrices could in turn provide a new level of support in the evolutionary dynamics of prebiological systems.